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
|
|---|---|---|---|---|---|
Dengue virus genotypes of Southeast Asian origin have been associated with higher virulence and transmission compared to other genotypes of serotype 2 ( DEN-2 ) . We tested the hypothesis that genetic differences in dengue viruses may result in differential binding to the midgut of the primary vector , Aedes aegypti , resulting in increased transmission or vectorial capacity . Two strains of each of the four DEN-2 genotypes ( Southeast Asian , American , Indian , and West African ) were tested to determine their binding affinity for mosquito midguts from two distinct populations ( Tapachula , Chiapas , Mexico and McAllen , Texas , USA ) . Our previous studies demonstrated that Southeast Asian viruses disseminated up to 65-fold more rapidly in Ae . aegypti from Texas and were therefore more likely to be transmitted to humans . Results shown here demonstrate that viruses from all four genotypes bind to midguts at the same rate , in a titer-dependent manner . In addition , we show population differences when comparing binding affinity for DEN-2 between the Tapachula and McAllen mosquito colonies . If midgut binding potential is the same for all DEN-2 viruses , then viral replication differences in these tissues and throughout the mosquito can thus probably explain the significant differences in dissemination and vector competence . These conclusions differ from the established paradigms to explain mosquito barriers to infection , dissemination , and transmission .
Dengue viruses , which cause millions of cases of dengue fever ( DF ) and dengue hemorrhagic fever ( DHF ) each year in over 100 countries , are transmitted by two species of mosquitoes , Aedes aegypti and Aedes albopictus . These vectors directly determine the rates of transmission of dengue viruses to humans , and they therefore determine the global spread and occurrence of disease . Thus , understanding the factors that make mosquitoes susceptible to infection by dengue viruses could help us design specific methods of control , whether these methods affect the mosquitoes directly ( e . g . genetic manipulation ) or indirectly ( e . g . population dynamics ) . The factors that influence the mosquito's capability for virus transmission , or vector competence , are numerous and have been reviewed in detail [1] , [2] . Vector competence is defined as the intrinsic permissiveness of a vector to infection , replication , and transmission of a virus . Specific factors , including mosquito and viral genetics and the environment , that govern Ae . aegypti transmission of dengue viruses ( members of the Flavivirus family ) , were reviewed recently [3]; however , it is important to point out that other flaviviruses , such as yellow fever and West Nile differ in their interactions with this mosquito . Gubler and Rosen were the first to study dengue vector competence , in both Aedes species , by comparing growth of representatives of the four different serotypes [4] , [5]; they were the first to describe a “gut barrier” for dengue virus dissemination in these mosquitoes . That is , dengue virus serotypes differed in their ability to escape two anatomical barriers to infection , the midgut infection barrier and the midgut escape barrier . The midgut barrier involves the ability of a virus to infect and replicate in midgut cells , with a variation in the ability to bind cell surface receptors or these cells not replicating virus ( non-permissiveness ) . The midgut escape barrier hinders the ability of virus to exit and disseminate to other tissues despite viral replication in the midgut , even to high titers . A barrier to virus transmission in the salivary glands was described in 1976 [6] and involves infection of those cells with no , or undetectable , virus being released in the saliva when mosquitoes bite . In contrast with most studies of vector competence , we have focused on comparing infection , replication , and transmission rates of dengue viruses in field-collected mosquitoes , with generations from eggs not higher than F4 ( i . e . , not lab-adapted mosquito colonies ) , and by comparing many different , low-passage ( i . e . , isolated from human patients or sylvatic mosquitoes , not lab-adapted ) virus strains of the same serotype ( DEN-2 ) . In addition , most of the viruses we used in our studies have been fully characterized , by determining their nucleotide sequences ( mostly entire genome or envelope gene only ) , by measuring their growth in primary human cells ( monocyte and dendritic cells , the natural targets of infection in human skin ) [7] , and by comparing their replication and disease causation ( virulence ) in a mouse model of dengue fever [8] . Thus , we have compared Ae . aegypti vector competence for a very broad spectrum of genetic variants belonging to one serotype of dengue , and with defined differences in phenotype ( growth , virulence , and epidemiologic ) and genotype ( phylogenetic grouping ) . In the experiments described here , we compared mosquito ex-vivo midgut binding of viruses belonging to the four genotypes of DEN-2 [9] from females of two mosquito colonies from different locations ( Tapachula , Chiapas , Mexico and McAllen , Texas , USA ) . Our results , in conjunction with those we reported previously on vast differences in virus infection and dissemination rates in these mosquitoes [10] , lead us to conclude that the midgut infection barrier for dengue viruses in Aedes aegypti may not be due to viral binding to these cells but rather to differences in viral genome replication ability . That is , differences in mosquito transmission abilities are probably determined by viral genetics and these determinants must be included in any attempts to control dengue virus transmission .
Eight DEN-2 virus strains , representing the Southeast Asian , American , Indian , and West African genotypes , were used in this study ( Table 1 ) . All viruses were low passage ( eight or less ) isolates from either patients or mosquitoes in order to reduce mutations associated with culture adaptation . Virus stocks were prepared as follows: the Aedes albopictus larval-derived cell line , C6/36 , was grown to 90% confluency in 175-cm2 flasks using growth media ( minimal essential medium , 10% fetal bovine serum [FBS , Atlanta Biologicals] , 2 mM L-glutamine , 1× non-essential amino acids , 100 u/mL of penicillin , and 100 µg/mL of streptomycin ) , then inoculated with dengue virus at a multiplicity of infection of approximately 10 genome equivalents/cell in 2 mL of maintenance media ( minimal essential medium , 2% FBS , 2 mM L-glutamine , 1× non-essential amino acids , 100 u/mL of penicillin , and 100 µg/mL of streptomycin ) and incubated for 1 hour at 28°C in an atmosphere of 5% CO2 . Flasks were then supplemented with 30 mL of maintenance media and maintained at 28°C in an atmosphere of 5% CO2 . Infection was monitored daily after 7 days by an indirect fluorescent antibody test ( IFAT ) and cell supernatants were harvested when more than 90% of the cells expressed dengue viral antigen , no more than 9 days . Individual virus aliquots were stabilized by the addition of 2% Prionex Reagent ( Calbiochem ) and stored at −70°C . Sterilized glass coverslips were inserted in each well of 6-well plates and seeded with 2×106 C6/36 cells with 2 mL of growth media . After two days , media was removed and replaced with serial dilutions ( 102–106 genome equivalents ) of virus in a total volume of 500 µL of maintenance media and incubated for 1 hour at 28°C in an atmosphere of 5% CO2 . An additional 2 mL of maintenance media was added and plates were incubated for 7 days at 28°C in 5% CO2 . An IFAT test was then performed to calculate the strain-specific TCID50 . Immunofluorescence was used to test for infection of C6/36 cells . For viral stock preparation , C6/36 cells were prepared for IFAT by spotting 10 µL of cell suspension onto a multi-well slide and incubated in a humid chamber for 20 minutes at 37°C . Cells were fixed by immersion in ice-cold acetone for 10 minutes . Slides were air-dried and each well was incubated with 20 µL mouse ascitic fluid containing anti-DEN-2 monoclonal antibodies ( MAb 3H5 , Centers for Disease Control and Prevention ) diluted 1∶200 in phosphate-buffered saline ( PBS ) for 30 minutes at 37°C . Slides were washed twice in PBS followed by incubation of each well using fluorescein isothiocyanate-labeled , goat anti-mouse IgG ( Sigma ) diluted 1∶200 in PBS for 30 minutes at 37°C . Slides were washed twice in PBS and immediately examined at 200× using a Nikon ( Melville ) Eclipse E400 microscope with an epi-fluorescence attachment . To determine viral titer by TCID50 the following modifications to the above protocol were made: glass coverslips were removed from six-well slides , immediately fixed in acetone and air dried . Coverslips were attached to glass slides using super glue ( Gorilla ) and rubber cement ( Elmer's ) was used to create wells on each coverslip . Aedes aegypti mosquito eggs were collected from Tapachula , Chiapas , Mexico and McAllen , Texas , USA during the spring and summer of 2010 to establish colonies . The F2–F3 ( Tapachula ) and F4 ( McAllen ) generations were used in this study . Mosquitoes were maintained in an insectary at 26–28°C , a relative humidity of 70–80% , and a 12∶12 hour light-dark cycle . Larvae were hatched and reared in pans of water at a density of 100–300 larvae per liter and fed a mixture of ground rabbit chow ( Purina ) ∶liver powder ( Bio-Serv ) ∶yeast ( Bio-Serv ) ( 4∶2∶1 ) ad libitum . Pupae were transferred to screened cages and emergent adults were maintained on an ad libitum diet of 10% sucrose ( Sigma ) in water . Successive Tapachula generations were generated by providing female mosquitoes a 37°C defibrinated rabbit-blood ( Colorado Serum Company ) meal using a water-jacketed membrane ( hog intestine ) feeder . Eggs were collected , kept moist for at least 24 hours , and air-dried prior to storage . In contrast to colony mosquitoes , mosquitoes used in experiments were not given blood meals but were maintained on 10% sucrose solution ad libitum . Midguts were dissected from 24 hour-starved adult ( 1–2 week old ) female Ae . aegypti in cold PBS+5% FBS [11] . Each replicate consisted of two intact midguts , with only the ends cut as a result of dissection with forceps . Midgut pairs were then washed once prior to incubation with virus and centrifuged for 3 minutes to aid in the supernatant removal . Pairs of midguts were incubated with 106–108 genome equivalents ( approx . 3 , 000–300 , 000 PFU total ) of virus in PBS+5% FBS in a total volume of 100 µL at 4°C for one hour . These virus quantities range from what a mosquito might ingest when biting a single viremic human to well above the estimated range ( see below ) . To remove unbound virus from midguts , the supernatant was removed after centrifuging for 3 minutes; 1 mL of fresh 4°C PBS+5% FBS was then added and transferred along with the midguts to a clean 1 . 5 mL microcentrifuge tube . A total of eight washes were necessary to remove virus not associated with the midguts , before the final RNA extraction . Total RNA was extracted from 50 µL of viral stocks and midgut tissue using Trizol reagent ( Invitrogen ) , following the manufacturer's instructions . Pelleted RNA was resuspended in a final volume of 50 µL of DEPC-H2O . Viral RNA copies in viral stocks and midgut tissue were estimated using a previously reported protocol [12] with some modification . RNA template ( 10 µL ) was amplified in duplicate using an RNA Ultrasense one-step quantitative RT-PCR system ( Invitrogen ) in a final reaction volume of 25 µL containing 1 . 25 µL of enzyme mix , 5 µL of 5× Ultrasense reaction mix buffer , 0 . 5 µL of 5′-carboxy-X-rhodamine reference dye , 5 µM of d2C16A primer ( 5′-GCTGAAACGCGAGAGAAACC-3′ ) [forward] , 5 µM of d2C46B primer ( 5′-CAGTTTTAITGGTCCTCGTCCCT-3′ ) [reverse] and 5 µM of VICd2C38B probe ( FAM-5′-AGCATTCCAAGTGAGAATCTCTTTGTCAGCTGT-3′-TAMRA ) . Amplification was performed on a 7500 real-time PCR system ( Applied Biosystems ) as follows: one cycle at 48°C for 30 minutes , one cycle at 95°C for 15 minutes and 40 cycles at 95°C for 15 seconds and 60°C for 1 minute . To estimate RNA copy number , a standard curve was generated using in vitro-transcribed RNA standards , as described below . To directly compare dengue viral copy number in ex-vivo infected midguts , the housekeeping gene Rp17S was used to quantify midgut tissue following a previously reported protocol [10] . RNA template ( 10 µL ) was amplified in duplicate following the RNA Ultrasense one-step quantitative RT-PCR system described above with the following modifications: The Rp17S specific forward primer ( 5′-ACATCTGATGAAGCGCCTGC-3′ ) , reverse primer ( 5′-ACACTTCCGGCACGTAGTTGT-3′ ) , and probe ( TET-5′-CACTCCCAGGTCCGTGGTATCTCCATC-3′-TAMRA ) replaced dengue specific primers and probe . Viral and Rp17S RNA standards were produced as follows: a 94 base pair fragment of the DEN-2 ( strain K0049 ) and a 101 base pair fragment of total mosquito ( Ae . aegypti ) RNA , were amplified by RT-PCR and separately cloned into the pCR2 . 1 plasmid using TOPO cloning kit ( Invitrogen ) . The recombinant plasmids were linearized with HindIII , and RNA transcripts were generated with a T7 Megascript kit ( Ambion ) following the manufacturer's instructions . Concentrations of transcribed RNAs were determined with a Ribogreen RNA quantitation kit ( Molecular Probes ) , and 10-fold serial dilutions were prepared and used to construct a standard curve . For comparison between virus genotypes and mosquito populations we used STATA v . 10 to perform a two-way factorial analysis of covariance ( ANCOVA ) with interaction . There were 35 mid-gut sample pairs for the American genotype , 36 Indian , 35 Southeast Asian , and 29 West African . There were 69 mid-gut sample pairs for the Tapachula and 66 for the McAllen populations . The dependent variable , number of virus genomes that bound to the midguts , was log10 transformed . The independent variables included the log10 transformed standardized viral genome equivalents added as the covariate and the four genotypes and two mosquito populations , both categorical . TCID50 s were calculated using the ID50 program , version 5 . 0 ( John L . Spouge , National Center for Biotechnology Information ) .
We found no statistically significant difference among the four DEN-2 genotypes in midgut binding affinity ( F ( 3 , 127 ) = 0 . 79 , p = 0 . 501 ) . This indicates that the number of virus genomes that bound to the midgut is not dependent on the viral genotype . The lack of significant differences among genotypes suggests there are no genotype-specific midgut receptors . Moreover , with the range of virus tested , within and exceeding what can be expected when mosquitoes bite a single viremic human ( approx . range: 0 . 1–1 , 000 PFU per µl ) [13]–[17] , the virus does not saturate available binding sites . Instead there is a linear relationship between the number of viral genomes that bind to the midguts and the number of viral genomes added to midguts ( Figure 1 ) . When controlling for the standardized viral genome equivalents added , we found a statistically significant difference between the two mosquito populations with respect to the DEN-2 genotypes we tested ( F ( 1 , 131 ) = 6 . 03 , p = 0 . 015 ) . When adjusting for the standardized viral genome equivalents added , the Tapachula mosquitoes bound more virus than mosquitoes from McAllen ( Figure 2 ) . This suggests that mosquitoes from the Tapachula site may be somewhat more susceptible to DEN-2 virus infection than mosquitoes from McAllen . Though the midguts were exposed to similar viral dilutions , the relationship between genomic equivalents ( RNA ) and infectious particles may vary among the genotypes . To determine if this was the case , TCID50 s were calculated for strains belonging to the four genotypes . Two genotypes , Southeast Asian ( K0049 TCID50 = 99 genomic equivalents ) and American ( IQT2913 TCID50 = 150 genomic equivalents ) , were found to generate or require a log or fewer genomic equivalents than the West African ( PM33974 TCID50 = 1110 genomic equivalents ) and Indian ( ArA6894 TCID50 = 2240 genomic equivalents ) genotypes , for infection of C6/36 cells . Differing TCID50 s between genotypes indicates a lack of consistency in the correlation between RNA transcripts measured by RT-PCR and infectious particles , a phenomenon we described previously [18] , as there is also a difference in plaque forming ability between DEN-2 strains and genotypes . However , we continue to use RNA equivalents as the best measure of viral quantities for standardization of inputs , and the testing of several dilutions per strain allows us to see this effect , independent of the actual number of RNA strands used for infection . As can be seen in Table 1 , we used a range of low passage number dengue virus isolates for these studies . The West African genotype viruses had the highest passage ( 4 and 8 ) , including replication in suckling mouse brain , whole mosquitoes and cell lines; the Indian genotype viruses were next highest ( 6 and 7 ) , including similar passage histories . The other two genotypes had the lowest number of passages and were limited to mosquitoes and cell lines . Because we used a range of virus inputs to infect dissected midguts , the trends of binding affinities shown in Fig . 1 reflected no differences between virus genotypes . Thus , there were no statistically significant differences in midgut binding that could be associated with virus passage levels .
In 2006 we reported significant differences in replication and dissemination of DEN-2 viruses of the American and Southeast Asian genotypes in Ae . aegypti from Texas [10] . The Southeast Asian genotype viruses had been shown to be associated with epidemics of DHF , while the American genotype viruses had not ( typically only DF ) ; the Southeast Asian viruses produced significantly more virus in ex vivo infected human dendritic cells , thus probably producing much higher viremias in patients; and these viruses grew and produced significantly higher disease signs in humanized mice , our animal model of DF [8] . Therefore , these characteristics seemed to explain the ecological or epidemiological displacement of the American genotype by the Southeast Asian genotype viruses in this continent and the displacement of the West African and Indian genotypes in those continents ( reviewed in [19] ) . However , the mechanism underlying these differences remained under debate . Our data reinforce the assertion that the observed differences between genotypes are the result of replication differences and are not due to a variation in midgut tissue affinity . Southeast Asian genotype viruses ( 3 strains ) outcompete American ( 3 strains ) genotype viruses in human dendritic cells and Ae . aegypti mosquitoes [20] . Viral replication begins within hours of infection [20] , [21] and in this study midguts were incubated with virus for 1 hour . Using orally infected mosquitoes from McAllen and Iquitos , Peru , Armstrong and Rico-Hesse [18] demonstrated that Southeast Asian DEN-2 viruses infected and disseminated much more efficiently than American viruses , when monitored over a two week time course . Similarly , Anderson and Rico-Hesse found the proportion of midguts from orally infected McAllen mosquitoes positive for DEN-2 antigen by IFAT over a two week time course were consistently higher for Southeast Asian viruses compared to American viruses , as were the titers based on quantitative RT-PCR [10] . Southeast Asian strains showed increased disseminated infection and susceptibility in mosquitoes from McAllen and Tehuantepec , Mexico , when compared to American genotype viruses [22] . A possible mechanistic explanation for these differences was provided recently using West Nile virus , a related Flavivirus , where virus diversification is driven by the mosquito RNA interference ( RNAi ) pathway [23] . The RNAi pathway is not activated until virus replicates within cells , a point which is beyond virus binding to cells . The lack of significant differences in DEN-2 genotype binding , as we show here , suggests that little selective pressure is placed on virus infection , in contrast to replication . Our focus on the first step in infection , virus binding to tissue , is unique and likely helps explain our findings which indicate no infection differences between genotypes . Instead , most researchers have looked at replication or dissemination as endpoint for infection [4] , [7] , [24] , [25] , [26] . Bosio et al [25] suggested a midgut infection barrier to explain dengue infection differences between mosquito populations . We find no evidence for an infection barrier based on the results described here . This is not the first time we have found evidence to implicate viral replication over binding affinity in defining infection rates . Though not specifically designed to assess whether virus binding was altered , our previous work showed that recombinant Southeast Asian genotypes lost their increased replication capacity in human cells when the Southeast Asian elements on the envelope glycoprotein ( E390 ) and 5′- and 3′- untranslated regions were replaced by American genotype structures [7] . We detected a significant difference when comparing the DEN-2 infection rates between the two mosquito populations . Finding differences in mosquito population responses to DEN-2 is consistent with those of Gubler et al . [4] and Black et al . , as summarized in a 2002 review article [3] . The latter found that mosquitoes from the Pacific coast of Mexico ( 57% ) were more competent as vectors for one strain of DEN-2 ( Jamaica 1409 ) than those from northeastern Mexico and Texas ( 53% ) . Likewise , Lambrechts et al . [24] highlight a significant interaction between virus genotype and mosquito population probably explained by co-adaptation . Though we did not see a virus genotype-specific interaction in our investigation of virus binding , we did measure mosquito susceptibility differences . Alternatively , although we did not measure size , the Tapachula mosquitoes appeared smaller than the McAllen mosquitoes in our colonies . Smaller sized Ae . aegypti females are more likely to become infected and disseminate virus than larger individuals ( based on wing length ) [27] . It may be that size differences account for differences in dose response . However , to truly test for mosquito population differences in midgut binding we would need to compare multiple DEN-2 strains binding across more than two mosquito populations . Our results are consistent with those of one midgut infection study done with genetically-selected Ae . aegypti strains , with varying susceptibility to dengue viruses [28] . Two low passage DEN-2 strains ( PR-159 , an American genotype , and SLK-1592 , an Indian genotype ) were used to infect two mosquito strains ( D2S3 and D2MEB ) in order to compare midgut replication and virus dissemination rates . Both mosquito strains were described as possessing high midgut infection rates but D2MEB has a lower dissemination rate compared to D2S3 when challenged with a high passage DEN-2 strain ( Jamaica , 1409 ) . Though the authors highlighted their findings using high passage DEN-2 viruses , their low passage viruses showed no differences between the two mosquito populations . That is , when utilizing a more natural experimental protocol , i . e . low passage virus , their findings were similar to ours of no midgut barrier . Unfortunately , all other studies of genetically-controlled determinants of vector competence done by this group of researchers have used only a high passage strain of DEN-2 ( Jamaica , 1409 ) [25] , [29] , [30] , and their results are therefore not comparable to ours . Thus , although we are using dissected midguts , with attempts to maintain the basal/luminal architecture , our results are consistent with other observations on virus replication and dissemination in the entire mosquito , including our own . In summary , we demonstrate that mosquito dissemination and transmission differences observed between DEN-2 genotypes are the result of viral replication and are not due to differences in viral binding to midgut cells .
|
Several factors , such as mosquito and virus genetics and environmental variables , determine the ability of mosquitoes to transmit dengue viruses . In this report , we describe new and important information that in some ways contradicts what is in the literature . Midgut infection barriers have been described as important determinants of virus transmission in mosquitoes but we found that virus binding to these midgut cells does not vary . When we compared binding of 8 different , low passage dengue viruses to mosquito midguts that were dissected out of Aedes aegypti mosquitoes ( the main vectors of dengue ) from Mexico and Texas , we found that there were no differences . Previously , we ( and others ) had shown that these same viruses differed significantly in replication and dissemination throughout the rest of the mosquito body , including the salivary glands , and therefore they differed greatly in their potential to be transmitted to humans . Thus , the data presented here are important considerations for future studies of vector competence and in determining strategies for control of dengue viruses in the vector .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"viral",
"transmission",
"and",
"infection",
"vector",
"biology",
"virology",
"emerging",
"viral",
"diseases",
"biology",
"microbiology",
"viral",
"replication",
"viral",
"vectors"
] |
2011
|
Variation in Vector Competence for Dengue Viruses Does Not Depend on Mosquito Midgut Binding Affinity
|
The intracellular endosymbiotic bacterium Wolbachia can protect insects against viral infection , and is being introduced into mosquito populations in the wild to block the transmission of arboviruses that infect humans and are a major public health concern . To investigate the mechanisms underlying this antiviral protection , we have developed a new model system combining Wolbachia-infected Drosophila melanogaster cell culture with the model mosquito-borne Semliki Forest virus ( SFV; Togaviridae , Alphavirus ) . Wolbachia provides strong antiviral protection rapidly after infection , suggesting that an early stage post-infection is being blocked . Wolbachia does appear to have major effects on events distinct from entry , assembly or exit as it inhibits the replication of an SFV replicon transfected into the cells . Furthermore , it causes a far greater reduction in the expression of proteins from the 3´ open reading frame than the 5´ non-structural protein open reading frame , indicating that it is blocking the replication of viral RNA . Further to this separation of the replicase proteins and viral RNA in transreplication assays shows that uncoupling of viral RNA and replicase proteins does not overcome Wolbachia’s antiviral activity . This further suggests that replicative processes are disrupted , such as translation or replication , by Wolbachia infection . This may occur by Wolbachia mounting an active antiviral response , but the virus did not cause any transcriptional response by the bacterium , suggesting that this is not the case . Host microRNAs ( miRNAs ) have been implicated in protection , but again we found that host cell miRNA expression was unaffected by the bacterium and neither do our findings suggest any involvement of the antiviral siRNA pathway . We conclude that Wolbachia may directly interfere with early events in virus replication such as translation of incoming viral RNA or RNA transcription , and this likely involves an intrinsic ( as opposed to an induced ) mechanism .
Arthropod-borne viruses ( arboviruses ) pose a considerable threat to human and animal health , yet effective control measures have proven difficult to implement [1 , 2] . In recent years novel means of reducing their replication in arthropod vectors have been suggested as an alternative way to reduce the prevalence of these viruses . One of the most exciting approaches is the use of the endosymbiotic intracellular bacterium Wolbachia to control arbovirus transmission from mosquito to vertebrate from within the arthropod vector [3 , 4] . Wolbachia was first found to confer resistance to viruses in Drosophila melanogaster [5 , 6] . When it was transferred to the mosquito Aedes aegypti it made the mosquitoes resistant to two important human pathogenic arboviruses , dengue virus ( DENV ) and chikungunya virus ( CHIKV ) [7 , 8] . Importantly , Wolbachia can also invade and be stably maintained in natural populations thanks to a trait called cytoplasmic incompatibility , which causes embryos to die when uninfected females mate with infected males [9] . This allows Wolbachia to spread through mosquito populations by providing a reproductive advantage to the Wolbachia-infected females that transmit the bacterium [10] . Field trials have shown that releasing Wolbachia-infected mosquitoes allows the bacterium to invade Ae . aegypti populations [11 , 12] and reduces the susceptibility of the mosquitoes to DENV [13] . The mechanism ( s ) by which Wolbachia confers broad resistance remains unclear . Antiviral protection is seen in insects that harbour high densities of Wolbachia [14 , 15] . For example Martinez et al ( 2014 ) showed a clear correspondence between Wolbachia density and the level of protection against the insect viruses , Drosophila C virus ( DCV ) and Flock House virus ( FHV ) [16] . This phenomenon is also seen in the mosquito Ae . albopictus , where the endogenous Wolbachia strains wAlbA and wAlbB have a relatively low density especially in key tissues such as the midgut and offer little protection against DENV [17 , 18] . It has also been hypothesised that Wolbachia protection is dependent on target cells and tissues harbouring Wolbachia [8 , 14 , 17] . Indeed , there is little evidence of Wolbachia and virus being present together in the same cell when either is present in a high density , suggesting that antiviral protection is cell autonomous [8 , 19] . It may be a case of competition for space or cellular resources [8] . Viruses and Wolbachia depend on host lipids , and in D . melanogaster it has been shown that enriching dietary cholesterol reduced the extent to which Wolbachia protects against DCV [20] . It has also been suggested that there is competition for iron resources within cells , as Wolbachia upregulates transferrin in mosquitoes while DENV and CHIKV are thought to cause its downregulation [21 , 22] . Viral replication is controlled by innate immune responses in both D . melanogaster and mosquitoes and several experiments suggested that the upregulation of immune pathways—immune-priming—may be important for Wolbachia-mediated antiviral activity [17 , 23 , 24] . However , this appears to only be the case in mosquito populations that have been transinfected with Wolbachia strains [21 , 24] . Drosophila species that are naturally infected with Wolbachia do not show an immune-priming phenotype , yet still confer antiviral activity [6 , 15 , 21 , 25] . RNAi is considered the most important antiviral response in insects , with double stranded viral RNA ( dsRNA ) being processed into short RNAs by the small interfering RNA ( siRNA ) pathway and directing the destruction of viral RNA [26 , 27] . However , several studies have shown that Wolbachia provides protection in mutant Drosophila and cells that lack components of this pathway , ruling out a role for the siRNA pathway in Wolbachia-mediated protection [19 , 28 , 29] . There is however data that suggest the miRNA pathway may play a role in Wolbachia mediated protection [30–32] . Wolbachia has been shown to alter the expression of multiple miRNAs in mosquitoes [31] . The miRNA pathway is involved in many cellular processes , and miRNAs are produced from genome-encoded nuclear precursor RNAs that are processed into 22 nucleotide ( nt ) molecules that can induce target RNA degradation or inhibition of translation [26 , 33] . In summary , the mechanism ( s ) by which Wolbachia confers antiviral activity are still unclear , and very little is known about exactly how the viral replication cycle is affected . Furthermore , it is not clear if Wolbachia itself responds to viral infection . In order to address these questions and to understand how Wolbachia interacts with viruses we have combined two powerful and well-studied model systems–the mosquito-borne alphavirus Semliki Forest virus ( SFV; Togaviridae , genus Alphavirus ) and a Wolbachia-containing D . melanogaster cell line [34]–and show that the endosymbiont provides strong protection against infection in this system . To identify the stage of the viral replication cycle that is likely being affected we compared SFV , SFV replicon and a SFV-based transreplicase system . We then used high-throughput sequencing to unravel the role of host small RNA pathways and the Wolbachia transcriptional response in antiviral protection . We find that Wolbachia targets the virus rapidly after infection , and is likely blocking early events in the replication of viral RNA ( for example translation of incoming RNA , the switch from translation to replication or RNA transcription ) within cells , though we cannot rule out an effect on entry or exit . These effects are neither associated with a host small RNA response nor a transcriptional response by the endosymbiont , but mediated by intrinsic activities .
We developed a model arbovirus infection system based on SFV , for which excellent molecular tools including replicons and recombinant viruses are available and which we have used extensively to study arbovirus-arthropod interactions [35–37] , and the D . melanogaster-derived Jw18 cell line infected with the Wolbachia strain wMel [34] . As SFV does not naturally infect D . melanogaster , we first established if this virus is able to infect and replicate in Jw18 cells . We used SFV4 ( 3H ) -RLuc , which expresses Renilla luciferase , RLuc , from the non-structural open reading frame ( Fig 1A ) and has been used previously to study antiviral mechanisms in mosquito cells [36 , 37] . The cells were first antibiotic treated to generate a Wolbachia-free control cell line , which we refer to as Jw18Free cells . These were infected at a multiplicity of infection ( MOI ) of 20 and cells were lysed 4 , 8 , 12 and 24 hours post infection ( hpi ) and RLuc activity measured . Over a 24 hour period RLuc activity gradually increased , indicating that SFV4 ( 3H ) -RLuc can infect and replicate in Jw18Free cells ( Fig 2A ) . In order to rule out any effect of Wolbachia or SFV on cell growth we next compared the growth of cells Jw18Free and Jw18Wol cells either infected or not infected with SFV4 ( 3H ) -RLuc . There was no significant difference between any of the treatments observed , indicating that neither Wolbachia nor SFV4 ( 3H ) -RLuc infection affected cell growth ( Fig 2B ) . It is important to note that SFV does not cause cytopathic effects in insect cells and therefore cells are able to continue to grow even under high infection rates . To determine if Wolbachia could protect against SFV infection in Jw18 cells , we infected Jw18Wol and Jw18Free cells with SFV4 ( 3H ) -RLuc ( infectivity >90% ) and measured RLuc activity at 7 and 24 hpi as a proxy for viral replication and spread [38] . Results indicated that even as early as 7 hpi inhibition of virus by Wolbachia is observed , with a 2–3 fold increase in RLuc activity in Jw18Free cells compared to Jw18Wol cells . By 24 hpi this difference is more marked with an 8–12 fold increase in RLuc activity in the Jw18Free cells ( Fig 2C ) . Therefore , Wolbachia confers antiviral protection against an arbovirus in this system . Furthermore the mechanism by which Wolbachia inhibits viral infection must be rapid suggesting either entry of the virus is inhibited or replication/translation are inhibited . As Wolbachia inhibits viral infection or subsequent processes as early as 7 hpi it could be hypothesised that entry of the virus into cells is inhibited , leading to a significant reduction in the number of subsequent replication complexes . In order to test this hypothesis we bypassed viral entry and analysed early translation and replication by transfection of in vitro transcribed capped SFV1 ( 3F ) RLuc-SG-FFLuc replicon RNA ( Fig 1B ) . In this SFV-derived replicon RNA , an open reading frame ( ORF1 ) encoding RLuc is fused to the non-structural nsP3 and the second , structural ORF ( 2 ) has been deleted and replaced with the FFLuc ORF ( see Materials and Methods ) . Alphavirus gene expression occurs in separate phases which are linked to replicative processes . Initially ORF1 is translated from the RNA genome giving rise to the nsP proteins , which carry out replicative functions . Then a switch from translation to replication occurs leading to production of a full length antisense copy of the genome , the antigenome . Antisense RNA ( which most likely exists in a duplex with the original positive-strand genome ) is used as a template for synthesis of new genomes; in addition it carries an internal promoter sequence that directs transcription of a subgenomic mRNA encoding the structural proteins ( Fig 1 ) [39] . Thus , expression of structural proteins ( or FFLuc marker , Fig 1B ) takes place only from these subgenomic RNAs i . e . is fully dependent on the replication process . In contrast , RLuc can be produced both by directly translating the replicon that was transfected into the cells as well as by translating new full-length positive strands , generated during RNA replication . Furthermore , in the absence of structural proteins no new virus particles can be formed preventing the spread of infection . It was found that Wolbachia results in a significant inhibition of early translation and/or replication independent of normal viral entry , with both RLuc and FFLuc readouts being significantly lower in Jw18Wol cells compared to Jw18Free cells ( Fig 3A and 3B ) . We cannot rule out that Wolbachia may also have an effect on entry which we do not observe in these assays . As this SFV-derived replicon allows for the separate analysis of transcription and translation from both the genomic and subgenomic promoters and corresponding mRNAs , it allows us to further pinpoint the stage in the replication cycle that is affected by Wolbachia . In the Wolbachia infected cells we observed a 200–600 fold decrease in FFLuc readout , a marker expressed from the RNA produced from subgenomic promoter , which is significantly greater ( T-Test P <0 . 0001 ) than the ~ 8 fold decrease in RLuc readout , a marker produced both from transfected RNAs and full-length positive-strand transcripts from the genomic promoter . This would suggest a clear and early inhibition of establishment of RNA replication . Alternatively , it could indicate a specific defect in the production of the subgenomic mRNA . This could occur either by Wolbachia directly interfering with replication , or by preventing the translation of proteins required for replication to occur and/or inhibiting the switch from replicase protein translation to RNA replication . Overall the results indicate that early viral RNA translation and/or replication were likely to be inhibited by Wolbachia . To further investigate the effect of Wolbachia on viral translation and/or replication we uncoupled viral replicase proteins from viral RNA by the introduction of two plasmids into cells in a SFV transreplicase assay . In this system , one plasmid encodes the viral replicase proteins and the second encodes an RNA template containing the untranslated regions of SFV with either FFLuc downstream of the genomic/actin promoter or Gluc downstream of the subgenomic promoter . In both cases the expressed sequences are under the control of a Drosophila Actin promoter . Upon expression the replicase proteins will bind the RNA template and replication of the reporter construct will take place . Expression of FFLuc is under both the control of the Actin promoter and the genomic promoter , therefore due to the high expression from the Actin promoter it is difficult to determine active replication from the genomic reporter . However active replication can be measured from the production of the Gluc reporter which is solely under the control of the subgenomic promoter . This system therefore allowed us to determine if production of replicase proteins from mRNA transcribed in the nucleus could overcome Wolbachia-mediated protection and determine if the origin of viral RNA is also important . In order to rule out arbitrary effects we also generated a non-replicating replicase with the introduction of a GDD-GAA mutation in the nsP4 protein , thus producing an inactive replicase . Results are shown in Fig 3C and 3D . Activity from the Actin and SFV genomic promoters appeared to be unaffected by the presence of Wolbachia , however it should be noted that there is no significant difference between the wildtype and mutant ( GAA ) replicase or cells where no replicase-expressing plasmid had been transfected . This confirms that the activity we saw here was most likely due to transcription from the Actin promoter and we are unlikely to detect expression from the genomic promoter as the system is at an optimum ( Fig 3C ) . This has also been observed in mammalian systems [40] . As shown in Fig 3D , we observed that Wolbachia significantly lowered activity from the subgenomic promoter as Gluc activity is lower in Jw18Wol cells compared to Jw18Free . This is in keeping with our previous observation that Wolbachia is able to inhibit viral translation and/or replication . Once viral RNA template is produced from the nucleus and transported to the cytoplasm , replication complexes are thought to form as normal . Thus Wolbachia was still able to confer protection even when viral replicase/RNA delivery routes were changed . Taken together this data strongly indicated that Wolbachia inhibits viral translation and/or replication . A major immune pathway in insects to fight viral infections is the exogenous siRNA pathway which involves the production of virus-derived small interfering RNAs ( viRNAs ) by the enzyme Dicer-2 acting on virus-derived dsRNA ( such as viral replication intermediates ) as a substrate [26 , 27 , 41] . The hallmark of this pathway in insects is the production of 21 nucleotide ( nt ) viRNAs , a process that has been described in detail for alphavirus infection of mosquitoes and mosquito cells [36 , 37 , 42] . To test whether the levels of viRNAs were affected by Wolbachia we used high-throughput sequencing of 18-33nt small RNAs from Jw18 cells 24hpi with SFV4 ( 3H ) -RLuc . In Jw18Free cells , small RNAs that were 21 nt long and map to the SFV genome were strongly induced upon viral infection ( Wilcoxon unpaired test: P = 0 . 008 , Fig 4A , S1A Fig for uninfected controls ) . 21 nt RNAs mapped equally in both sense and antisense orientations to the viral genome ( Fig 4A; P > 0 . 1 Chi-squared test against a uniform distribution ) . The length and lack of strand bias or first nucleotide bias ( Fig 4A ) suggest that these small RNAs are generated by the activity of Dicer-2 on virus-derived dsRNA , probably replication intermediates; moreover 21 nt viRNAs were distributed across the viral genome as previously reported for SFV ( Fig 5A ) [37 , 43] and also other arboviruses [26 , 27] . Indeed there was no nucleotide bias seen at any position either in the Jw18Free or Jw18Wol cells ( S2A and S2B Fig ) . If Wolbachia infection were to reduce viral infection by upregulating antiviral RNAi we would expect increased viRNA production in the presence of Wolbachia . However , whilst 21 nt viRNAs were still present above the background seen in virus-free controls ( P = 0 . 008 , Wilcoxon unpaired test; Fig 4B and S1B Fig ) , the amount of viRNAs was strongly reduced on both sense and antisense orientations relative to Wolbachia-free cells ( P = 0 . 008 , Wilcoxon unpaired test; Fig 4A versus 4B ) . Due to a significant reduction in viral replication , the distribution of 21 nt viRNAs from Jw18Wol cells revealed very few areas of viRNA production ( Fig 5B and 5C ) . RNAs smaller than 21 nt were similar between Jw18Wol and Jw18Free cells in infected cells ( Fig 4A and 4B ) , however it is likely that these smaller species of RNAs are background against the D . melanogaster genome or degradation products . This confirmed that Wolbachia does not protect against infection by enhancing the production of small RNAs against viruses . Instead , these results are consistent with a model whereby Wolbachia interferes with viral replication , leading to a decrease in the levels of viral replication intermediates and therefore a reduction in dsRNA , the substrate available for Dicer-2 and exogenous siRNA pathway induction . This is not surprising as previous studies have shown that Wolbachia can still confer antiviral activity in flies mutant for key components of the siRNA pathway [29] . Previous studies suggested that Wolbachia affects the sensitivity of mosquito cells to viral infection by altering host miRNAs levels [31 , 32] . Therefore , we tested whether Wolbachia alters the expression of known miRNAs in Jw18 cells . In the absence of virus , no miRNAs had significantly different expression levels in Jw18Wol and Jw18Free cells ( Fig 6 and S3A Fig ) . It is likely therefore that any differences are not important to Wolbachia mediated protection as in our system there are no significantly altered miRNAs between non-infected Jw18Free and Jw18Wol cells . We next examined whether Wolbachia alters the miRNA response to viral infection . We identified a number of miRNAs that significantly changed in abundance when Jw18Free cells were infected with SFV ( Figs 6 and S3D; S1 Table ) . Very similar changes in miRNA expression were seen when cells with Wolbachia were infected with virus ( Fig 6 and S3B Fig , S1 Table ) . However , when cells were infected with SFV there were no miRNAs whose abundance was significantly changed by the presence of Wolbachia ( Fig 6 and S3C Fig ) . Therefore in D . melanogaster cells there is no evidence that Wolbachia modulates the constitutive expression of miRNAs or the miRNA-mediated response to infection . The similar miRNA response to virus in cells with and without Wolbachia is intriguing as Wolbachia-infected cells have greatly reduced levels of virus , and suggests it may reflect a sensitive response to initial infection by the virus . To investigate whether Wolbachia itself might mount an active antiviral response after infection , we tested if there is a transcriptional response of Wolbachia to viral infection . We sequenced total RNA from Jw18Wol cells 7 and 24 hpi with SFV4 ( 3H ) -RLuc virus together with uninfected controls . Over 229 million reads could be mapped to the D . melanogaster , Wolbachia or SFV transcriptomes ( excluding D . melanogaster rRNA ) . Of these 85 . 6% mapped to D . melanogaster and 12 . 4% mapped to Wolbachia . In the virus-infected cells , 3 . 8% of reads mapped to the SFV genome , and this dropped to 0 . 03% for cells that were not challenged with the virus . No Wolbachia genes were differentially expressed in response to viral infection at either time point ( Fig 7B and 7C ) . There are three reasons to believe that this is a true lack of a transcriptional response and not simply a lack of statistical power . First , across many transcripts we were able to detect a transcriptional response of the cells to SFV in the same experiment ( Fig 7D and 7E ) , and the coverage of many of these differentially expressed transcripts was lower than for many Wolbachia transcripts ( Fig 7B–7E ) . Second , we had a very large dataset . In each of the 4 treatments about two thirds of the Wolbachia transcripts had over 100 reads mapped to them ( Fig 7A ) , which is expected to give good power to detect differential expression . Compared to most published RNAseq experiments ours was a highly replicated experiment involving 40 libraries ( biological replicates ) across 4 lanes of Illumina HiSeq . Third , if genes with very low-expression are ignored , our estimates of Wolbachia gene expression levels in cells with and without SFV were nearly identical ( Fig 7B and 7C , log2FC≈0 ) . Therefore , the lack of differential expression cannot be attributed to a lack of statistical power .
The bacterial symbiont Wolbachia offers an exciting opportunity in the fight against arbovirus transmission by mosquitoes . Several studies have found that it has antiviral activity in key arbovirus mosquito vectors [8 , 28 , 44 , 45] . However , the exact mechanisms behind this activity are poorly understood . In order for Wolbachia to be used as a long term and sustainable system to control arbovirus spread , it is critical that we understand these mechanisms . By combining two powerful model systems–the model arbovirus SFV , and a Wolbachia-infected D . melanogaster cell line–we were able to show that Wolbachia may protect against virus at a very early stage of infection and appeared to block replication and/or translation of viral RNA . This did not involve an active transcriptional response from either the host , the small RNA pathways or Wolbachia itself . Little is known about how Wolbachia affects viral replication and dynamics , with studies measuring changes in either the survival of infected insects or viral copy number [5 , 6 , 17–19] . To investigate the phenomenon in more depth we used the alphavirus SFV , for which virus-encoded reporter genes are known to correspond well to viral replication [38] . We found that Wolbachia is able to inhibit viral replication as early as 7 hpi . To our knowledge this is the first indication that Wolbachia inhibits viral replication at such an early stage ( for example by inhibition of initial translation of incoming RNA or other early replicative processes ) , as previous studies have focused mainly on days post-infection [5 , 17 , 19 , 46] . This suggests that the mechanism by which Wolbachia is conferring antiviral activity is either fast acting or is already present upon viral infection i . e . intrinsic . This is important as for Wolbachia to be used successfully as a control mechanism for arbovirus transmission , control of viral replicative processes at an early stage could be vital , as it would not allow the virus to replicate to high levels allowing for dissemination . The SFV life cycle can be divided into entry , initial translation of incoming viral genomes , switch from translation to replication , RNA replication , translation of structural proteins from subgenomic RNA , assembly and exit from cells . Understanding where in this lifecycle Wolbachia acts is vital to understanding the mechanisms behind its antiviral activity . This can often prove difficult to investigate as little is understood about SFV ( and arbovirus ) replication in insect cells compared to vertebrate cells . By utilising SFV reporter viruses and replicon constructs we can begin to deconstruct the replication cycle . The very early stage at which SFV was inhibited and the fact that bypassing viral entry still resulted in viral inhibition by Wolbachia suggests that Wolbachia is likely blocking an early post-entry stage in the replication cycle . We can further dissect how Wolbachia affects the replication cycle due to the presence of a sub-genomic promoter in alphaviruses [39 , 47] . Expression from the first ORF of SFV does not require viral replication ( though due to an increase of genomic RNA copy number , it is dramatically increased when replication occurs ) , as positive-sense RNA virus genomes can act as mRNA immediately after infection . In contrast translation and protein expression from the second ORF of SFV requires a full round of replication in order for the sub-genomic promoter to become available . Therefore by utilising the replicon SFV expressing RLuc reporter directly from genomic RNA and FFLuc via activity of the subgenomic promoter we could test whether Wolbachia is affecting the replication of viral RNA . Our results showed that viral replication and/or translation of replicon were inhibited by Wolbachia . This could occur either by Wolbachia directly interfering with replication , or by inhibiting the translation of viral proteins required for this to occur and/or inhibiting the switch from translation to replication . Further to this we showed by using a SFV transreplicase system and uncoupling the replicase proteins from template RNA that there was significant inhibition of the subgenomic promoter and thus viral translation and/or replication . Thus providing replicase proteins either directly or separately did not overcome the inhibition phenotype . These results indicate that the origin of viral RNA ( transcription in the nucleus in the transreplicase system ) was also not important to Wolbachia-mediated antiviral activity . This block in translation/replication in the presence of the endosymbiont was backed by analysis of the induction of small RNA responses , which reveal that Wolbachia-infected cells did not show higher production of viRNAs which are derived from the dsRNA generated during viral replication ( see below ) . One possible way in which Wolbachia could protect against viruses would be by the bacterium mounting an active antiviral response following infection . If this were the case , it would likely be reflected in a transcriptional response of the endosymbiont to the virus . However , in a very large dataset we did not find a single Wolbachia gene that was differentially regulated in response to viral infection . This is consistent with the very early inhibition of viral replication , which suggests that the antiviral mechanism is constitutively present before the virus infects the cell and is thus intrinsic . Previous studies have indicated that a host miRNA response may be responsible for Wolbachia’s antiviral activity [30 , 31 , 48] , but we found no support for this hypothesis in our system . In virus-free mosquitoes Wolbachia changes the expression of a number of miRNAs , but we found no such differences in our cell line ( S3E Fig ) . Previous studies did not analyze concurrent infections of both Wolbachia and virus [30 , 31 , 48] . When we did this we found a marked miRNA response to the virus , but this was unaltered by the presence of Wolbachia . It is surprising that we still see host miRNA response to viral infection in the presence of Wolbachia , as the level of SFV infection is extremely low . This suggests that miRNA response may be due to early events in viral replication such as virion binding and/or entry or that this response requires very little viral protein synthesis/RNA replication to be initiated . Previous studies have indicated that the viRNA pathway is not required for Wolbachia’s antiviral activity [29] . Our data supports this , with a strong reduction in 21 nt viRNAs mapping to the SFV genome and antigenome when Wolbachia is present . This suggests that viral replication is inhibited so significantly that very few viRNAs are produced , rather than Wolbachia inducing an antiviral RNAi response . Further studies utilising this system would be beneficial to the field . For example little is known about the possibility of viruses mutating to overcome Wolbachia mediated protection as long term studies are lacking . A cell-based assay offers an ideal opportunity to look at virus evolution over the long term in such associations . In addition to this it would be interesting to look at the effect on other viruses in our system as other studies have indicated that even within the same family Wolbachia can have different effects on viruses [49] . In Drosophila studies Wolbachia is also known to protect against FHV without lowering viral titres . It would be interesting to look at this in the context of our findings , as it may suggest another mechanism by which Wolbachia can confer antiviral activity [6] . In conclusion we have developed a powerful new system to study the replication dynamics of SFV in Wolbachia-infected D . melanogaster cells . While the exact mechanism of the antiviral response remains unknown , current data is consistent with a ‘passive’ mechanism such as competition for resources or space , or by Wolbachia remodelling the intracellular environment . While effects of Wolbachia on entry or exit cannot be excluded , our data point to an effect on translation and/or replication at least for this model alphavirus . Considering the broad antiviral effects of Wolbachia across Drosophila and mosquito species , it is tempting to propose a model of inhibition that relies on similar intrinsic mechanisms rather than diverse processes such as miRNA regulation or immune responses . The data presented in this study point towards such an antiviral mode of action by Wolbachia endosymbionts .
The Wolbachia-infected D . melanogaster cell line Jw18Wol ( obtained from W . Sullivan , L . Serbus , A . Debec ) has been described elsewhere [34] . A corresponding tetracycline cured line ( Jw18Free ) was created by treating cells with 10 μg/ml of tetracycline for 4 passages , cells were then tested for Wolbachia by PCR and DAPI staining and if the infection was cleared cells were left for 5 more passages in order to eliminate tetracycline effects . Cells were maintained at 24°C in Shields and Sanger media ( Sigma ) supplemented with 10% fetal calf serum ( FCS ) and 10% Penicillin/Streptomycin ( Pen/Strep ) . Cells were checked every four passages for the presence of Wolbachia by PCR using two separate primer pairs as described previously [6]; cells were stained with DAPI in order to visualise Wolbachia content ( shown as dots in cytoplasm ) , density in Jw18Wol cells was consistent with previous observations with ~90% of cells infected ( S4 Fig ) . qPCR was also carried out in order to determine Wolbachia density . Standard curve analysis was carried out and normalised to an Actin endogenous control . Primers used were as follows , Actin5CF_ GACGAAGAAGTTGCTGCTCTGGTTG Actin5CR TGAGGATACCACGCTTGCTCTGC and WolF GTTTGCAATACAACGGTGAA WolR CAACCCTGATGTCGTCCATT . qPCR was carried out using the ABI Fast SYBR Green Master Mix , as per manufactures guidelines , on an ABI 750 Fast machine . Results indicated that when compared to Actin endogenous control there is ~22 times more Wolbachia ( S5 Fig ) , suggesting a density of at least 100% with more than one copy number per cell as is seen in the DAPi staining . SFV4 ( 3H ) -RLuc virus was grown and cultivated as described previously [37] . For replicon production , pSFV1 ( 3F ) RLuc-SG-FFLuc plasmid ( details available on request ) was linearized with SpeI and purified using the PCR product purification kit ( Roche ) . 1 μg of linearized DNA was in vitro transcribed using MEGAscript SP6 polymerase kit ( Ambion ) in the presence of cap analog ( Ambion ) . pAc51-V5-His backbone was used to construct the plasmid expressing the replicase of SFV . First , the multiple coning site of the plasmid was replaced by a polylinker sequence TTCGAATATGGATCCTATTTAATTAATATCCTAGG ( recognition sites of Bsp119I and PacI are underlined ) . Second , Bsp119I and PacI adapters were added to the 5’ and 3’ ends of the sequence encoding SFV replicase , respectively , by using PCR and subcloning procedures . Finally , the sequence encoding SFV replicase was inserted into modified pAc51 vector using Bsp119I and PacI restriction sites . The resulting plasmid was designated as pAct51-SFVRepl . In order to obtain plasmid encoding for a polymerase defective form of SFV replicase , the conserved GDD motif in nsP4 was altered to GAA using PCR-based mutagenesis and subcloning procedures; the resulting plasmid was designated pAc51-SFVRepl-GAA . In order to obtain a plasmid for expression of template RNA for the SFV replicase the polylinker and terminator regions of pAc51-V5-His were replaced with a synthetic DNA fragment consisting from the first 280 nucleotides of SFV genome ( including EcoRV restriction site , nucleotides 275–280 ) followed by the sequence TATGGATCCTATGGCGCGCCGTCGAC ( recognition sites BamHI , BssH2 and SalI underlined ) . The replacement was performed in such a way that the 5’ end of the SFV genome was positioned exactly downstream of last start site of actin promoter . The following sequences were added using synthetic DNA fragments ( GenScript , USA ) and subcloning procedures: a ) sequence encoding for firefly luciferase ( FFLuc ) , placed in frame with N-terminal fragment of SFV nsP1 ( amino acid residues 1–65 , encoded by the 5’ region of SFV genome ) ; b ) SFV subgenomic promoter spanning from position -145 to +51 with respect of the start site of SFV subgenomic RNA; c ) sequence encoding for Gaussia luciferase ( Gluc ) ; d ) 3’ UTR of SFV followed by 69 adenine residues; e ) negative strand ribozyme of hepatitis delta virus; e ) late termination region of simian virus 40 . Elements a , b , c and d were separated from each other by recognition sequences of EcoRV , ApaI and Bsp119I nucleases , respectively . The plasmid was designated as pAc51-Temp-Fluc2-SG-Gluc . Cells were plated out at a density of 3 . 5x105 cells/well 24 h prior to infection in a 24 well plate . For infection , virus was diluted in Shields and Sanger media ( Sigma ) with no FCS . Virus was titred as described [37] and an MOI of 20 was shown to give an infectivity of over 90% in Jw18Free cells ( S4C Fig ) . Cells were incubated for 3 h before media was replaced with fresh media supplemented with 10% FCS and Pen/Strep . Samples were collected at time points as indicated . Virus free cells were mock infected by treating the same as infected cells but without the addition of virus to the media . Wolbachia was detected using the nuclear stain DAPI . Briefly cells were fixed in 4% paraformaldehyde , permeabilized and covered in Vectashield containing DAPI . Wolbachia was indicated by the presence staining outside of the nucleus which was absent in tetracycline treated cells . SFV was detected as described previously using an antibody against the replicative protein nsP2 [50] . Cells were plated out at a density of 3 . 5x105 cells/well 24 h prior to transfection in a 24 well plate . Cells were transfected with 1 μl of in vitro transcribed RNA using Fugene in Shields and Sanger media minus FCS . Cells were incubated for 2 h before media was replaced with fresh media supplemented with 10% FCS and Pen/Strep . For transreplicase experiments 300 ng of each plasmid was transfected into cells as described above . Cells were lysed in passive lysis buffer and luciferase readings carried out using the Dual Luciferase Kit ( Promega ) . Luciferase activities were determined on a Glomax Multi+ Microplate Multimode reader ( Promega ) . Cells were infected with SFV as described above . At 7 and 24 hpi cells were lysed in 1 ml of Trizol solution . Small RNA libraries were prepared according to the Illumina method using the Truseq small RNA Preparation kit . We made a small adjustment to the manufacturer’s protocol to include a ribosomal RNA blocking step prior to ligation of 5´ adapter and reverse transcription in order to eliminate the abundant 30 nt D . melanogaster 2S rRNA , as specified in [51] . Small RNA libraries were then sequenced on one lane of a HiSeq 2000 . High throughput sequencing data was processed and aligned to the viral genome as described previously [52] . miRNAs were mapped to D . melanogaster miRNAs downloaded from miRBase [53] using a custom Perl script and analysed to identify statistically significant changes in expression according to the negative binomial distribution using the R package DESeq as described [54] . All data processing was carried out in the R statistical environment . Sequence data has been submitted to the Sequence Read Archive ( SRA ) under accession number PRJEB9710 . Wolbachia-positive Jw18Wol cells were infected with SFV at an MOI of 20 as described above . 7 and 24 hpi cells were lysed in 1 ml of Trizol; total RNA was extracted using the Direct-Zol MiniPrep kit ( Zymo ) . The extracted RNA was then treated with TURBO DNase ( Ambion ) and purified using the RNA Clean and Concentrator kit ( Zymo ) . D . melanogaster ribosomal RNA was then depleted using the Ribo-Zero Gold Magnetic kit ( Human/Mouse/Rat , Epicentre ) . Libraries of the rRNA-depleted total RNA were prepared at The Genome Analysis Center ( Norwich ) with the Truseq RNA Sample Preparation kit ( Illumina ) and sequenced in 2 lanes on a HiSeq 2000 . Sequence data has been submitted to the Sequence Read Archive ( SRA ) under accession number PRJEB10681 . Sequences were quality trimmed from the 5´ and 3´ ends using Trimmomatic version 0 . 30 [55] when average quality scores in sliding windows of 4 base pairs dropped below 20 or when the quality score at the beginning or end of the read dropped below 20 . Sequences less than 25 bp in length were discarded . Reads were aligned to the transcriptomes of Wolbachia strain wMel ( Genbank accession number: GCA_000008025 . 1 ) and D . melanogaster ( BDGP v . 5 . 25 ) , and to the genome of SFV ( Genbank: KP699763 . 1 ) . Alignments were performed using Bowtie2 version 2 . 1 . 0 [56] with default parameters , and splicing was allowed in D . melanogaster using TopHat2 version 2 . 0 . 9 [57] with default parameters and no novel junctions allowed . The numbers of reads per transcript were counted using HTSeq [58] for Wolbachia and D . melanogaster . Differential expression analysis was performed using edgeR [59] . Lowly expressed genes were filtered out by requiring that each gene have at least 1 count per million in at least 8 samples . Differential expression in response to viral infection was measured separately at 7 and 24 hpi . Significance was assessed using exact tests [59] with a FDR of 10% .
|
The intracellular endosymbiotic bacterium Wolbachia can protect insects against viral infection . However , the mechanisms underlying this antiviral activity are poorly understood . We have developed a new model system combining Wolbachia-infected Drosophila melanogaster cell culture and the model mosquito-borne virus , Semliki Forest virus . Wolbachia confers strong antiviral activity against SFV . Our study indicates that viral replication appears to be inhibited at a very early stage , such as initial translation or replication . Results indicate that Wolbachia does not mount a transcriptional response to SFV infection and that host small RNA pathways are not involved in Wolbachia mediated antiviral activity in our system . We conclude that Wolbachia may directly interfere with early events in virus replication such as translation of incoming viral RNA or RNA transcription , and this likely involves an intrinsic ( as opposed to an induced ) mechanism .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"microbiology",
"animals",
"wolbachia",
"invertebrate",
"genomics",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"micrornas",
"bacteria",
"drosophila",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"infectious",
"diseases",
"gene",
"expression",
"arboviral",
"infections",
"viral",
"replication",
"insects",
"animal",
"genomics",
"arthropoda",
"biochemistry",
"rna",
"nucleic",
"acids",
"virology",
"protein",
"translation",
"genetics",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"genomics",
"non-coding",
"rna",
"organisms"
] |
2016
|
Wolbachia Blocks Viral Genome Replication Early in Infection without a Transcriptional Response by the Endosymbiont or Host Small RNA Pathways
|
Understanding the relationship between topology and dynamics of transcriptional regulatory networks in mammalian cells is essential to elucidate the biology of complex regulatory and signaling pathways . Here , we characterised , via a synthetic biology approach , a transcriptional positive feedback loop ( PFL ) by generating a clonal population of mammalian cells ( CHO ) carrying a stable integration of the construct . The PFL network consists of the Tetracycline-controlled transactivator ( tTA ) , whose expression is regulated by a tTA responsive promoter ( CMV-TET ) , thus giving rise to a positive feedback . The same CMV-TET promoter drives also the expression of a destabilised yellow fluorescent protein ( d2EYFP ) , thus the dynamic behaviour can be followed by time-lapse microscopy . The PFL network was compared to an engineered version of the network lacking the positive feedback loop ( NOPFL ) , by expressing the tTA mRNA from a constitutive promoter . Doxycycline was used to repress tTA activation ( switch off ) , and the resulting changes in fluorescence intensity for both the PFL and NOPFL networks were followed for up to 43 h . We observed a striking difference in the dynamics of the PFL and NOPFL networks . Using non-linear dynamical models , able to recapitulate experimental observations , we demonstrated a link between network topology and network dynamics . Namely , transcriptional positive autoregulation can significantly slow down the “switch off” times , as comparared to the nonautoregulatated system . Doxycycline concentration can modulate the response times of the PFL , whereas the NOPFL always switches off with the same dynamics . Moreover , the PFL can exhibit bistability for a range of Doxycycline concentrations . Since the PFL motif is often found in naturally occurring transcriptional and signaling pathways , we believe our work can be instrumental to characterise their behaviour .
Synthetic biology can be used to uncover the design principles of natural biological systems through the rational construction of simplified regulatory networks [1] . So far , it has been shown that feedback and feed-forward loops are essential regulatory motifs involved in transcriptional regulation [2] . For example , natural circadian clocks consist of feedback loops in which a set of transcriptional activators and repressors are connected by mutual feedback; these clocks are able to maintain a 24-h periodicity in protein expression [3] . How exactly this is achieved is still under debate , but it is an intrinsic property of the network topology [4] . Therefore , understanding the relationship between topology and dynamics of transcriptional regulatory networks in mammalian cells is essential to elucidate the biology of complex regulatory and signaling pathways . In this work , we aimed at characterising an inducible transcriptional positive feedback loop ( PFL ) in mammalian cells consisting of the CMV-TET promoter , responsive to the Tetracycline-controlled transactivator tTA , driving expression of the tTA protein itself and of a fluorescent reporter protein ( Fig . 1A ) . Most of the studies carried out so far in mammalian cells are based on plasmid transfection , which prevents precise quantitative measurements due to the unpredictable amount of plasmids that enters each cell , and to the transient nature of transfection . To overcome these limitations , we generated a clonal population of Chinese Hamster Ovary cells ( CHO ) carrying a stably integrated version of the PFL , which were then used to evaluate the dynamical properties of this transcriptional motif . We also performed control experiments by generating a clonal population of cells lacking the positive feedback loop ( NOPFL ) , as shown in Fig . 1B . In vivo quantitative measurements of fluorescence intensity in time , following addition of the inducer molecule ( Doxycyline ) able to”switch off” the network , were used to fit an Ordinary Differential Equations ( ODE ) model of the PFL and NOPFL networks . The models were able to reproduce the experimental data , and , interestingly , highlighted differences in the dynamic properties of the PFL versus the NOPFL networks , which are due to the intrinsic differences in the two network topologies . It has been suggested theoretically that the PFL motif can slow down response times of transcriptional regulatory networks [5] . Here , we experimentally demonstrated that this is the case , and we reported for the first time a quantitative characterisation of a transcriptional positive feedback loop in mammalian cells , which can be instrumental in better understanding the properties of natural occurring transcriptional and signaling networks .
Our approach is based on the use of well known and characterised regulators of gene expression , in order to achieve a complete control of the network behaviour . The positive feedback loop ( PFL ) is shown in Fig . 1A . In particular , we took advantage of the inducible Tet regulatory system; the expression of Tetracycline-controlled transactivator tTA is self-controlled by a CMV-TET promoter , responsive to the tTA itself unless the Tetracycline , or its analogous Doxycycline , is added to the medium in which cells are grown [6] . To follow the dynamics of the PFL , a destabilised yellow-green variant of the enhanced green fluorescent protein ( d2EYFP ) ( Clontech ) , with a reported half-life of approximately two hours , had to be placed under the control of the same promoter . To this end , we constructed a unique cassette with an Intra Ribosomal Entry Sequence ( IRES ) in between of the transactivator tTA and the d2EYFP , which enables a single mRNA to encode for two different proteins ( Fig . 1A ) . In order to stably express in CHO cells the PFL network , and to generate a clonal population , we inserted the cassette in Fig . 1A in a lentiviral vector [7][8] . Infected cells were first sorted by Fluorescence Activated Cell Sorter ( FACS ) and then a clonal population of CHO cells carrying the PFL was generated by single cell expansion ( PFL cells ) ( Experimental procedure: cell culture , lentiviral transduction , switch-off experiment ) . To capture the dynamic properties intrinsic to the PFL , we needed to generate a control network lacking the positive feedback loop ( NOPFL ) , but using the same biological “parts” as in the PFL network . As shown in Fig . 1B , we constructed a cassette containing the same CMV-TET promoter upstream of the d2EYFP . The tTA protein , this time , was placed under the control of a constitutive promoter , thus breaking the feedback loop . Using the same strategy described above , we generated a clonal population of CHO cells carrying the NOPFL network ( NOPFL cells ) . We experimentally verified that both PFL and NOPFL cells have the same number of tTA/d2EYFP DNA integrations ( Fig . 1 inset ) . ( Experimental procedure: DNA extraction , RealTime PCR ) . We experimentally evaluated the degradation rate of the reporter protein ( d2EYFP ) since this is a key parameter which affects the observed fluorescence dynamics . To evaluate d2EYFP degradation rate , stably integrated NOPFL cells were treated with Cyclohexamide to a final concentration of , , or , to inhibit protein synthesis [9] . The fluorescence intensity of NOFPL cells was followed for 12 hrs and images were acquired at 15 min intervals . The resulting d2EYFP dynamics are shown on Fig . 2 and appear very similar , independently of the Cyclohexamide concentrations . The experimental data were fitted to an exponential curve , and the degradation coefficient was used to obtain the half-life ( ) of the d2EYFP protein: = log ( 2 ) / ( Fig . 2 and Table 1 ) . We estimated to be in the range 3 . 6 h–4 . 4 h . ( Experimental procedure: determination of d2EYFP half-life ) . The estimated value is about two-fold the reported d2EYFP half-life of 2 h [10]; we believe that this discrepancy is likely due to the fact that cells were grown at a temperature , rather than the usual . To observe the dynamics of the PFL and NOPFL networks , we performed time-series experiments in which stably-integrated CHO-PFL cells and CHO-NOPFL cells were imaged using time-lapse fluorescence microscopy ( Experimental procedure: cell culture , lentiviral transduction , switch-off experiment ) . The experimental design consisted in treating both PFL and NOPFL cells with different amounts of Doxycycline in order to “switch off” the circuit , by preventing the tTA protein from binding the CMV-TET promoter . We tested the following Doxycyline concentrations: , , and and followed the dynamic behaviour of both the PFL and NOPFL cells for 43 h , collecting images every 15 min , and quantifying the average fluorescence intensity of the cell population ( Image acquisition and processing ) . In this way , we averaged out cell-to-cell variability in the response , since at the beginning of each experiment the tracked microscopy field contained at least 15 cells . Experiments were carried out at a temperature of in order to limit cell motility and reduce the risk associated to data loss occurring when cells exit the tracked field [11] . The average fluorescence intensity of the reporter gene across the cell population for both the PFL and NOPFL networks is shown in Fig . 3 for the different concentrations of Doxycycline indicated . In Fig . 4 replicate time-course experiments are shown for each of the Doxycycline concetrations used . The most striking feature is the slow down in the switch off time of the PFL cells as compared to the NOPFL cells; moreover the switch off time of the PFL is affected by Doxycycline concentrations , whereas NOPFL cells always switch off with approximately the same dynamics . We derived a model of the PFL and NOPFL networks using Ordinary Differential Equations ( ODEs ) . ODEs are commonly used to describe the average behaviour of a population of cells [12] and have been shown to be appropriate for the analysis of synthetic networks in a number of cases ( [13] , [14] , [15] , [16] , [17] , [18] ) . In our settings , the use of such a modelling approach is valid , since we are measuring the average behaviour of a clonal population of cells . For each species , i . e . each mRNA and correspondent protein concentration , we wrote one equation , which expresses the change in concentration of the species in a given time interval , as the result of a production term and a degradation term . We assumed: The last assumption was introduced in order to take into account d2EYFP maturation time needed for correct protein folding [16] . Thus , we introduced two differential equations as in [16]: one for the translation of mRNA to the unfolded d2EYFP protein , and one for the folded protein d2EYFP . Letting be the tTA IRES d2EYFP mRNA concentration , the tTA protein concentration , the unfolded d2EYFP protein concentration and the folded d2EYFP protein concentration , the PFL network can be described as follows: ( 1 ) ( 2 ) ( 3 ) ( 4 ) Note that , due to the presence of the IRES sequence , the concentrations of tTA protein and d2EYPF protein depend on the same variable ( ) , that is the concentration of the single mRNA transcript encoding for both proteins . For the NOPFL network , we let represent only the d2EYFP mRNA concentration , and we assumed a constant level ( ) of the tTA protein , due to the constitutive promoter driving expression in the NOPFL cells . The equations thus become: ( 5 ) ( 6 ) ( 7 ) The parameter estimation problem can be defined as an optimisation problem , where the goal is to minimise a performance measure defined as the error between model predictions and observations , which in our case are the time-series during the “switch off” experiments in Fig . 3 and 4 ( Model simulations and parameter identification ) . In our setting , there are 12 parameters to be fitted , 11 of which are common to both the PFL and NOPLF models ( Table 1 ) . Several alternative strategies can be pursued in order to obtain best estimates of the model parameters , ranging from Newton's method to Genetic Algorithms ( GAs ) . In this work we employed the Trust Region method ( TRM ) implemented in PottersWheel [19]; thanks to the multimodel and multiexperimental capabilities of this tool , we were able to identify the 12 parameters by simultaneously fitting Eqs . ( 1 ) to ( 7 ) to all of the experimental time-courses at once . These time-courses include all of the different Doxycycline concentration for both the PFL and NOPFL cells for a total of 24 time-courses , when taking experimental replicates into account . In order to estimate confidence intervals for each parameter , we run multiple times the TRM identification procedures on the same experimental time-series data , using parameter perturbation routines to allow extensive sampling in the parameters' space . In order not to introduce any specific bias in our search , we only set admissible ranges for the parameter values to be identified , which reflected physical and biological feasibility , either obtained from literature [15] or directly estimated ( degradation rate of = ) . The result of the parameter fitting procedure are reported in Table 1 along with the estimated standard deviation , which are in most of the cases one order of magnitude less than the corresponding parameter value , or at most of the same order of magnitude . We observed that the paramter in Table 1 , which affects the strength of Doxycycline repression on the tTA protein activity , is much smaller than 1 . Usually Hill coefficients are greater than 1 , therefore we wondered what could be a possible biological explanation for this behaviour . We observed that for the range of Doxycycline concetration used in the experiment ( to ) , using the parameters' values in Table 1 , the function: can be approximated by the function: ( and ) . This means that a Michealis-Menten function can also describe the effect of Doxycyline on tTA acitivity , but a certain level of leakiness ( ) must be taken into account; that is even for large concentrations of Doxycycline , the activity of the tTA protein cannot be completely shut down . The “switch off” time-series experiments were simulated with both the PFL and NOPFL models using the fitted parameters as shown in Fig . 3 . The inferred models are able to recapitulate the observed dynamics in response to different inducer concentrations and experimental settings . Observe that the parameters for both the PFL and NOPFL models are identical , except for in the NOPFL equations , which is not present in the PFL model . Hence , the observed differences in the dynamical behaviour of the PFL and NOPFL networks are due to the intrinsic differences in their topology , and are robust to changes in parameters values , as demonstrated in the next section . In order to further investigate the relationship between topology and dynamical properties , we first observed that the NOPFL model described by Eq . 5–7 is a system of linear time-invariant ODEs , for which the theory of liner dynamical systems applies [20] . From the theory , we know that changes in Doxycycline concentration in Eq . 5 will not affect the dynamic behaviour of the model , which is governed by the smallest among three degradation terms ( Model simulations and parameter identification ) . The concentration of Doxycycline affects only the steady-state values , i . e . how much the network will switch off , but not its dynamics , i . e . how fast it will switch off . Therefore , independently of the values of the parameters , the model of the NOPFL network predicts that for any concentration of Doxycycline , the network will switch off with the same dynamics , albeit possibly reaching different steady-state levels . Fig . 5 reports the “switch off” time , , for both the PFL ( dashed ) and the NOPFL ( solid ) networks as a function of Doxycycline concentration , computed via numerical simulations of the two models with the parameters estimated in Table 1 . is defined as the time taken by the fluorescence intensity to reach of its final steady-state value ( OFF ) , following treatment with Doxycycline at a given concentration ( Material and Methods ) . As expected , the for the NOPFL network is constant and does not change with Doxycyline . This is in agreement with the experimental observations; in Fig . 5 , the switch off time for the NOPFL network for the different concentration of Doxycycline was estimated from the experimental time-series data ( in Fig . 5 ) ( Model simulations and parameter identification ) . On the other hand , the PFL network has a very different behaviour , as can be seen in Fig . 5 . Specifically , for a range of Doxycycline concentrations , the PFL is considerably longer ( in Fig . 5 ) that the NOPFL counterpart , which again is in agreement with the experimentally observed behaviour ( Material and Methods ) . In order to investigate the origin of the observed dynamical behavior of the PFL circuit , we analysed the phase portrait associated to the d2EYFPtTA mRNA and the tTA protein , which allows to directly observe trajectories of two state variables at once . Moreover , by imposing the steady-state conditions ( i . e . ) , we can derive nullclines , as well as , the their intersection points , which correspond to the equilibrium points of the network . In Fig . 6 the nullclines for different Doxycycline concentrations are shown . When no Doxycycline is present , two stable points ( OFF and ON ) and one unstable equilibrium points coexist in the same phase portrait , thus providing evidence for the bistability of the PFL network , a shared property among positive feedback loops [21] . However , as Doxycycline concentration increases , the bistability is lost ( Fig . 6 ) , and the only possible equilibrium point is the “OFF” state .
We have demonstrated , in a mammalian experimental system , that a transcriptional positive feedback loop can slow down the “switch off” times , as compared to an equivalent network without autoregulation . The reason for a cell to “choose” a PFL control strategy for transcriptional regulation , rather than the NOPFL strategy , could be due to the intrinsic robustness of this approach to transient activation of the network . For example , in a signalling pathway , a ligand ( equivalent to Doxycycline in our PFL ) could cause a transcription factor to stop transcribing itself , as well as , a set of target genes , to initiate a specific response . However , in order for the pathway not to respond to a transient concentration of the ligand , the PFL strategy has to be chosen , otherwise the response would start immediately ( NOPFL case ) . Moreover , the response time of the PFL network can be modulated by the ligand concentration , if this is really high , the system will switch off as quickly as possible ( Fig . 5 ) , alternatively the ligand can be present at low , or medium , concentration , but it should persist for a long time , in order for the pathway to respond . This kind of behaviour has been recently described as “persistence detection” in cellular signal processing to indicate the ability of the genetic circuit to distinguish between transient and persistent signals [22] . Interestingly , it has been shown in E . coli [23] that the transcriptional negative feedback loop ( NFL ) has an opposite effect , that is , it can significantly speed up the rise-times of transcription , but has very little effect on the switch-off times . From numerical simulations , we verified that for the PFL the slow down effect is only in the switch off times , whereas rise-times are barely affected ( data not shown ) . The duality between positive and negative feedback has been predicted in a biological setting [5] , and it is a well established concept in “control engineering” , a branch of engineering which deals with the design of automated mechanisms to control a variable of interest ( the altitude of an airplane , or more simply , the temperature of a room via thermostat ) [24] . Specifically , the negative feedback loop is a classic control engineering approach to speed up the response times of a sytem , thus quickly achieving a desired value of a variable of interest . Positive feedback loops , instead , can slow down the response of the system to external input , and are used by control engineers to build “memory” elements , also known as switches , which are able to be in one of two steady-states ( OFF or ON ) , and which are robust against unwanted transient perturbations that may inadvertently switch off ( or on ) the system . We indeed verified that the PFL can exhibit bistability for zero or low concentrations of Doxycyline ( Fig . 6 ) . A bistable genetic network will cause a population of cells to divide in two sub-populations , each in one of the two possible states ( OFF or ON ) . In yeast , this has been experimentally verified using a simple PFL based on the rtTA system [21] . In our mammalian PFL , this behaviour was not detected experimentally ( FACS analysis data now shown ) . This can be easily explained by observing that the PFL model is bistable but the basin of attraction of the OFF equilibrium point ( Fig . 6 ) is much smaller as compared to that of the ON state , when no Doxycycline is present . Therefore , just few cells will be in the OFF state and these will not be enough to be significantly detected experimentally . We predict however that for intermediate concentration of Doxycycline ( in Fig . 6 ) the basin of attraction will be comparable and bistability should be detected experimentally . We believe our work can be instrumental to characterise the behaviour of naturally occurring regulatory pathways in mammalian cells and to prove that , despite the complexity of these pathways , they may be understood relying on the knowledge of the behaviour of simplified regulatory network as the one here reported .
To implement the gene circuit in a lentiviral vector , we used the ViraPower Promoterless Lentiviral Gateway Expression System ( Invitrogen ) which takes advantage of the site-specific recombination properties of bacteriophage lambda , making the transfer of single DNA sequences faster than the usual cloning strategies . The pMAtTA-IRES-EGFP vector containing the transactivator tTA , the IRES element and the enhanced green fluorescent protein ( EGFP ) was synthesised by GENEART together with the recombination sites . The d2EYFP was amplified from pd2EYFP-1 ( Clontech ) by PCR with a forward primer containing a NheI recognition sequence ( 5′-CATGGCTAGCATGGTGAGCAAGGGCGAGGAG-3′ ) and a reverse primer containing an EcoRV recognition sequence ( 5′- ATTCGATATCAGTCGCGGCCGCATCTACA-3′ ) . The PCR product and pMAtTA-IRES-EGFP were then digested with NheI-EcoRV restriction enzymes and the d2EYFP ligated in place of EGFP , generating a new vector termed pMAtTA-IRES-d2EYFP . The pMAtTA-IRES-d2EYFP was then linearised with the AseI restriction enzyme and recombined with the pDONR221 ( Invitrogen ) following the manufacturer instruction . In this way we generated pENTRtTA-IRES-d2EYFP vector with specific recombination sites . The CMV-TET promoter was amplified from pTRE2 ( Clontech ) by PCR . The PCR was performed with the Taq polymerase provided by Invitrogen that adds a single deoxyadenosine ( A ) to the 3′ ends of PCR products . This allows PCR inserts to ligate efficiently with the pENTR5′-TOPO vector which is supplied linearised with single 3′-deoxythymidine ( T ) overhangs , obtaining the pENTR5′-TOPO-CMV-TET with specific recombination sites . Finally we performed a recombination reaction between the pENTRtTA-IRES-d2EYFP , pENTR5′-TOPO-CMV-TET and the pLenti/R4R2/V5-DEST according to manufacturer instructions . To generate the lentiviral vector containing the gene expression cassette lacking the positive feedback loop ( NOPFL ) , the d2EYFP was amplified from pd2EYFP-1 with the High-Fidelity Taq Phusion ( Fynnzimes ) which gives blunted-end PCR product . The forward primer ( 5′-CACCGCCACCATGGTGAGCAAGGGCGAGGAG-3′ ) was designed to allow the direct cloning in pENTR directional TOPO vector ( Invitrogen ) , generating the pENTR d2EYFP vector . Then we performed a recombination reaction between the pENTR d2EYFP , pENTR5′-TOPO-CMV-TET and the pLenti/R4R2/V5-DEST according to manufacturer instructions . As suggested by the manufacturer , the lentivirus was then produced in 293FTcells . 293FT cells were maintained at in a 5% CO2-humidified incubator , and cultured in DMEM ( GIBCO BRL ) supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) ( Invitrogen ) , 1% L-glutamine , 1% MEM Non-Essential Amino Acids , 1% MEM Sodium pyruvate and 1% antibiotic/antimycotic solution ( GIBCO BRL ) . CHO cells were maintained at in a 5% CO2-humidified incubator , and cultured in -MEM ( Sigma ) supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) ( Invitrogen ) , 1% L-glutammine and 1% antibiotic/antimycotic solution ( GIBCO BRL ) . CHO AA8 TET-OFF cell line ( Clontech ) was maintained -MEM ( Sigma ) supplemented with 10% TET system approved FBS ( Invitrogen ) , 4 mM L-glutamine , G418 ( Sigma ) , and 1% antibiotic/antimycotic solution ( GIBCO BRL ) . To transduce cells with the virus produced , 500 , 000 CHO and CHO AA8 TET-OFF cells were plated and incubated overnight . On the day of transduction the medium was removed and 1 mL of the virus was added to the cells together with polybrene ( Invitrogen ) to a final concentration of 6 ug/mL . After an overnight incubation the medium containing the virus was removed and replaced with complete culture medium containing Blasticidin ( Sigma ) to a final concentration of to select for stably transduced cells . Cells were sorted for fluorescence intensity using a BD FACSAria Cell Sorting System ( Becton Dickinson ) . d2EYFP was excited at 488 nm , and emission was detected using a 525 nm bandpass filter . Serial dilutions of stably transduced cells ( up to 0 . 05 cells/mL ) were plated in 96-well microtitre plates , and dilutions containing only one cell per well were selected . Monoclonal colonies were cultured and amplified as described , to obtain monoclonal populations . For the switch off experiment , 500 stably-integrated-CHO and CHO AA8 TET-OFF cells were plated in chamber slide ( lab-Tek ) and treated with Doxycycline ( Clontech ) to a final concentration ranging from to ) . The switch off experiments were repeated twice for the and Doxycycline concentrations , while 3 and 5 replicates were obtained for and . Experiments were performed in parallel for both the PFL and NOPFL cells . 1 , 000 , 000 PFL and NOPFL cells were plated in a 6-well multiwell plate to reach a confluence of 80 at the moment of the DNA extraction . The day after cells were collected and resuspended in of PBS after centrifugation for five minutes at 300×g . Then the DNA was extracted using the DNeasy Blood and Tissue kit ( Qiagen ) . We compared the DNA levels of tTA and d2EYFP in NOPFL cells and PFL cells by RealTime PCR following DNA extraction , proving that the both cell populations carry a unique copy of the networks in their genome . Quantitative RealTime PCR reaction were set up in duplicates using the LightCycler 480 SYBR green master mix ( Roche ) and the amplification was performed using a LightCycler 480 RealTime PCR instrument ( Roche ) . The PCR were carried out using the following primers: d2EYFP forward ( 5′-acgacggcactcaagacc-3′ ) ; d2EYFP reverse ( 5′-gtcctccttgaagtcgatgc-3′ ) ; PFL tTA forward ( 5′-aaagcagctgaagtgcgagag-3′ ) ; PFL tTA reverse ( 5′-gatggtgctgccgtagttgtt-3′ ) ; NO PFL tTA forward ( 5′-acagcgcattagagctgctt-3′ ) ; NO PFL tTA reverse ( 5′-acctagcttctgggcgagtt-3′ ) . Data analyses were performed using the LightCycler 480 Software ( Roche ) . GAPDH DNA levels were used to normalise the amount of DNA and Cts were calculated as the difference between the average GAPDH Ct and the average tTA and d2EYFP . Images were acquired using an inverted epifluorescence microscope ( Nikon Eclipse TI-E , Nikon Instruments ) equipped with an incubation chamber ( H201-OP R2 , Okolab ) , a digital camera ( Andor Clara , Andor ) , a 20× objective ( Obj . CFI PF DLL 20× Ph1 , Nikon ) and a 512-nm/529-nm ( B/G/R ) d2EYFP-specific excitation/emission filter set . Temperature was maintained at a constant level as the experimental setup required , while CO2 concentration was set to be 5% of the total air volume injected in the incubation chamber . Both phase-contrast images and fluorescent fields were acquired at intervals of 15 minutes . Exposure times for the phase-contrast field was set to ( transmitted light lamp voltage was set to ) while ( Intensilight lamp set at of the maximum power ) was chosen as exposure time for the fluorescent images: this choice was meant to prevent photobleaching while optimising the ratio between the quality of the images and reflected-light-induced stress on the cells . Experiments were carried out using NIS-Elements AR v . 3 . 10 644 ( Nikon Instruments ) software package and the Perfect Focus System ( Nikon Instruments ) to maintain the same focal plane during the whole duration of the experiment . At the end of the acquisition process , images were extracted as raw data for the fluorescence quantification procedures . The experiments were set up so that at the beginning of each experiment the first image contained at least 15 cells and no more than 30 cells , to avoid cells exiting the image during the time-lapse experiment due to cell replication and “over-crowding” . Image segmentation was carried out in Mathworks Matlab R2010b ( Mathworks Inc . ) ; the algorithm we implemented to quantify fluorescence was meant to distinguish the foreground ( living cells ) from the background in each image of the bright field . We used morphological operators such as erosion and dilation ( imerode and imdilate functions from the MATLAB image processing toolbox ) . Thus two binary masks were built in order to compute separately the mean d2EYFP fluorescence of the foreground and the background using an element by element matrix multiplication between the binary images and the fluorescent one . The average fluorescence intensity across the cell population was then computed as the difference between the foreground and the background for each image at each time point ( i . e . no single cell fluorescence quantification is performed ) . To evaluate d2EYFP degradation rate , 500 stably integrated CHO tetOFF cells were plated in chamber slide ( lab-Tek ) and , after cell adhesion , Cyclohexamide ( Sigma , stock dilution 10 mg/ml in sterile water ) was added to the medium to a final concentration of , , or 500 . Temperature was maintained at . Image acquisition and analysis was performed as described above . The experimental data were fitted into an exponential curve using the curve fitting tool ( cftool ) from Matlab 2010b , and the degradation coefficient was used to obtain the half-life ( ) of the d2EYFP protein: = log ( 2 ) / Numerical simulations were run using Matlab 2010b ( Mathworks Inc . ) . We used ode23s solver ( a detailed discussion of the numerical methods used by ode23 can be found in [25] ) . For the parameter identification , we used the PottersWheel toolbox [19] implemented in MATLAB . Two sets of parameters were identified: the dynamical parameters governing the model and a scaling factor meant to approximate the transduction contribution of the microscopy equipment . Since Doxycycline has been only added at time min in our experiment we forced the fitting procedures to start from the model predicted ON steady state . We defined the following objective function: ( 8 ) where is the number of experimental data points , are the predicted values of the mathematical model ( using the inferred parameters ) , are the experimental data points and is the sample variance computed over the experimental replicates . As optimisation algorithm we used Trust Region Method ( TRM ) in a logarithmic parameter space: at the iteration of the optimisation procedure the TRM approximates the shape of the function to be minimised with the model thus trying to solve the following problem: ( 9 ) being the new parameter vector considered as solution at the iteration . If the model has quadratic form the vector can be obtained by observing that: ( 10 ) and therefore ( 11 ) In order to allow an extensive exploration of the parameters' space , and to avoid local minima , we used a quasi-random number generator routine in PottersWheel [19] to select an initial guess of the parameters' values , and then launched the TRM procedure M times ( M = 100 in our settings ) , requiring the cost function in eq . 8 to be [19] . The values in Table 1 represent parameters for which the cost function ( eq . 8 ) is the smallest across the M runs; whereas the standard devation of each parameter in Table 1 is evaluated by considering all of the M runs . Moreover , in order to compare switch off times among the different experiments , we computed the defined as the time the circuit needed to achieve the of the mean initial fluorescence calculated for each experiment as follows: ( 12 ) with fluorescence of the frame in the sequence smoothed by moving average filtering .
|
Synthetic Biology aims at designing and building new biological functions in living organisms . At the same time , Synthetic Biology approaches can be used to uncover the design principles of natural biological systems through the rational construction of simplified regulatory networks . Mathematical models of the networks are then derived from physical considerations and can be used to explain the observed dynamical behaviours . We have characterised a regulatory motif often found in transcriptional and signalling pathways . We constructed a positive feedback loop motif in mammalian cells , consisting of a protein controlling its own expression . We have shown that this motif exhibits a dynamic behaviour which is very different from that obtained when the autoregulation is removed . This difference is intrinsic to the specific wiring diagram chosen by the cell to control its behaviour ( feedback versus non-feedback configurations ) , and can be instrumental in understanding the complex network of regulation occurring in a cell .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"bioengineering",
"systems",
"biology",
"biological",
"systems",
"engineering",
"control",
"theory",
"mathematics",
"molecular",
"biology",
"applied",
"mathematics",
"synthetic",
"biology",
"biology",
"nonlinear",
"dynamics",
"genetics",
"and",
"genomics",
"engineering"
] |
2011
|
Construction and Modelling of an Inducible Positive Feedback Loop Stably Integrated in a Mammalian Cell-Line
|
Why is spatial tuning in auditory cortex weak , even though location is important to object recognition in natural settings ? This question continues to vex neuroscientists focused on linking physiological results to auditory perception . Here we show that the spatial locations of simultaneous , competing sound sources dramatically influence how well neural spike trains recorded from the zebra finch field L ( an analog of mammalian primary auditory cortex ) encode source identity . We find that the location of a birdsong played in quiet has little effect on the fidelity of the neural encoding of the song . However , when the song is presented along with a masker , spatial effects are pronounced . For each spatial configuration , a subset of neurons encodes song identity more robustly than others . As a result , competing sources from different locations dominate responses of different neural subpopulations , helping to separate neural responses into independent representations . These results help elucidate how cortical processing exploits spatial information to provide a substrate for selective spatial auditory attention .
Past studies of spatial effects in auditory cortex have focused on how spatial location is encoded . These studies typically find that single-unit spatial tuning in cortex is weak [1]–[4] , not topographically organized [5] , [6] , and not encoded independently of other perceptually important features [7] . There is good evidence for a specialized “where” pathway in auditory cortex , in which spatial information plays a larger role than in other cortical areas [8] . However , although we know of no single study that directly compares spatial tuning in cortex to that of lower stages of the auditory pathway , spatial tuning of cortical neurons is generally broad compared to both behavioral sensitivity [4] and spatial encoding in the midbrain [9] , [10] . One hint for how to resolve these apparent discrepancies is that in an awake animal performing a spatial task , spatial information in cortical responses is enhanced [10] . Together , these results suggest that although spatial information is available , it is not the primary feature represented in the cortical auditory regions . Instead , spatial information may modulate neural responses in a way that depends on task demands , thus enabling analysis of sound sources in realistic auditory scenes [11] , [12] . It may be that spatial effects are not best revealed by looking at how well source location is encoded by neural responses , but rather by examining how source location affects other aspects of information in cortical spike trains . In everyday perception , source location matters most in auditory scenes in which sounds compete with each other . Although listeners can localize a sound source in quiet , this ability is degraded in more typical , real-world settings containing reverberant energy or competing sources [13] . In contrast , in exactly those kinds of realistic situations where there are competing sources , spatial separation helps listeners segregate sounds and enables them to focus selective attention , a critical skill for understanding a source of interest [14] , [15] . In this sense , behavioral results support the idea that the locations of competing sources strongly influence auditory perception , regardless of whether the listener can effectively localize in such a setting . Motivated by these observations , we hypothesized that the effects of spatial location on cortical processing would best be revealed by a study that uses competing sound sources . Rather than focusing on how accurately spatial location of a source was encoded , we explored how competing source locations influenced the ability to encode the identity of a target communication signal ( in this case , birdsong ) . We found that , consistent with our hypothesis , source location of a target song presented in isolation had little effect on how well neurons in avian field L ( the analog of mammalian primary auditory cortex [16] ) encoded song identity; however , in the presence of a competing noise masker , both target and masker locations strongly influenced encoding of song identity . Moreover , depending on the location of target and masker , different neurons were “best” at encoding identity . Such a coding scheme may provide a substrate for spatial auditory attention , as top-down modulatory control signals could selectively suppress responses of neurons favoring a masker in order to reduce competition and allow more precise analysis of a target from a desired location .
We recorded neural responses from male zebra finches in the auditory forebrain ( field L , based on stereotactic coordinates [17]–[19] ) to stimuli from four azimuthal locations in the frontal hemifield . Target stimuli were two conspecific songs , presented either in quiet ( “clean”; Figure 1A ) or in the presence of a spectrally similar noise masker coming from the same or a different location as the target song ( Figure 1B ) . We assessed neural performance using a single-trial spike-distance-based [20] nearest-neighbor classification scheme [21] , calculating a percent correct score that indicates how well neural responses coded stimulus identity . Chance performance was 50% . Consistent with prior studies [17] , [22]–[24] , rate coding alone was insufficient to allow reliable stimulus discrimination; mean performance when no masker was present was only 54% , averaged across recording sites . In each experimental session , there were four loudspeaker locations , leading to 16 target-masker spatial configurations . If the recording electrode was in the left hemisphere , loudspeaker locations were on the left side ( −90° , ipsilateral to the electrode ) , in front ( 0° ) , halfway between front and right ( +45° ) , and on the right ( +90° , contralateral ) . These locations were flipped about the midline when recording in the right hemisphere . Henceforth , coordinates are referenced to the recording electrode , so that ipsilateral azimuths have negative signs and contralateral azimuths have positive signs . Discrimination performance was calculated for all 16 configurations and three signal-to-noise ratios ( SNR; −6 dB , 0 dB , +6 dB ) . To assess the extent to which the head created an acoustical obstruction ( “head shadow” ) to the ear opposite the sound source , we measured sound level at both ears from all four locations using a masker token as the probe stimulus . The differences between left and right ears were 1 . 5 , 0 . 1 , −0 . 8 , and −1 . 3 dB for −90 , 0 , +45 , and +90° , respectively . For the example site in Figure 1A and B , clean performance was near ceiling at all tested locations . Masked performance was much lower and varied substantially as the target was moved from the ipsilateral side ( −90° ) to the contralateral side ( +90° ) , holding the masker at −90° . Across recording sites , the masked performance varied much more than clean performance did as a function of location . To quantify this , we computed the spatial sensitivity ( defined as the difference between the best and worst performance for a given experimental condition; see Materials and Methods ) for each site for both clean and masked targets . Spatial sensitivity was 3-fold higher with a masker present than without ( p< . 001; Figure 1D ) . The driven spike rate in response to clean songs did not vary significantly with location ( r = . 16 , p = . 068; Figure 1E ) . This distinction is important: while target azimuth was at best weakly coded by the rate response of the neurons , information about song identity encoded in spike trains varied greatly with target and masker locations . The way in which classification performance varied with spatial configuration varied from site to site . Indeed , some sites responded best when the target was in a particular hemisphere ( Figure 2A , site 1 ) , some for a particular target-masker location configuration ( sites 2 and 3 ) , and some in idiosyncratic configurations that fit no simple description ( site 4 ) . To explore how such a population of neurons might encode song identity , we considered two population-coding schemes . The first was based on a previous study , which assumed that behavioral performance was determined by the best thresholds across a population of neurons , an approach termed the “lower envelope” principle [25] . Here we define the corresponding neural “upper envelope” as the best classification performance across the entire neuronal population . The performance of individual sites and the upper envelope are shown in Figure 2B as a function of target and masker location for an SNR of −6 dB . While no one site performs well for all spatial configurations , almost all configurations yield at least some sites that encode target identity well . At higher SNRs , the upper envelope is at ceiling ( Figure 3A ) . To better reveal the effects of spatial configuration , we calculated the mean performance across sites for each spatial configuration . Despite the complex dependence of performance on spatial configuration for many of the sites , the mean performance varies smoothly with spatial configuration for each SNR . Specifically , mean performance is best when the target is contralateral and the masker ipsilateral to the neural recording site and worst in the reverse configuration ( Figure 3B ) . Figure 3C shows the mean performances across sites in which the target is farther than the masker from the recording site , in the contralateral direction . Representing the data this way assumes a simple population model in which the neurons in one hemisphere are favored over the other ( i . e . , the responses from the hemisphere contralateral to the target are enhanced and the ipsilateral responses are suppressed ) . Using this model ( which includes only the values in the lower right half of the grids in Figure 3B , including the diagonal ) , the effect of spatial separation ( as well as SNR ) is highly significant ( p< . 001 for both ) ; moreover , linear regression fits at each SNR show that performance improves with increasing spatial separation of target and masker . Such performance increases are parallel with results from behavioral studies in humans [26] and birds [27] that report spatial unmasking . Maskers degrade responses to target songs . A simple way to evaluate the masker interference is to compare the response elicited by the target in quiet to that of responses to the target plus masker . Differences between the two responses can be categorized into orthogonal categories of spike additions , where the presence of the masker causes extra spikes ( usually in the gaps between syllables ) , and spike subtractions , where spikes that are elicited by the target alone are reduced by the presence of the masker ( usually during syllables; see Figure 4A–C ) . Both types of interference have been studied before [18]; here we extended that analysis . We modeled spike trains that had only subtractions or only additions ( Figure 4D; see Materials and Methods ) , and then calculated performance for these modeled spike trains just as we did for the measured ones . We first validated our modeling approach by comparing predictions for modeled spike trains containing both additions and subtractions ( i . e . , the full effect of the masker ) to measured data ( see Figure 4 , “modeled” rasters and performances ) . The example model rasters look similar to the measured masked spike trains , and target song identification performance closely matched performance using the masked spike trains . These results validate our methods for modeling additions and subtractions . Following validation , we modeled spike trains that included only spike additions or only spike subtractions to separate their relative effects on performance . When modeling spike additions only ( i . e . , when no subtractions were modeled ) , target identification was better than for the measured response . On the other hand , performance for subtractions-only spike trains was only slightly better than the measured responses for two of the three configurations . For the target-contralateral , masker-ipsilateral configuration ( right column of Figure 4 ) , performance was essentially equal for the subtractions-only and masked spike trains . These results suggest that additions did not impair discrimination performance when the target was contralateral to the recorded site . However , including additions had some impact on the other two configurations . Overall , this analysis shows that the masker degraded performance more by preventing spikes that a clean target would have elicited than by causing additional spikes . The times at which spikes are likely to be added by the masker tend to occur when the clean response rates are low . This can be quantified by correlating the clean stimulus response rate ( Figure 4A ) with the rate of subtractions ( blue depths in Figure 4C ) as a function of time . This correlation is significant and negative , confirming that subtractions reduce spikes the most when the likelihood of a spike in response to the clean stimulus is great ( r = − . 75 , p< . 001 ) . In contrast , the correlation between the time-dependent spike additions ( red peaks in Figure 4C ) and the clean rate is weak ( r = . 08 , p< . 001 ) . Taken together , these results suggest that the effect of removing spikes from the peaks interferes with target identification more than adding spikes . This holds true even in spatial configurations where the number of spikes added is greater than the number of spikes removed .
Here we show that , in quiet , sound source location has only a modest impact on coding of song identity in field L , an analog of auditory cortex [16] . In general , spatial tuning in brainstem is sharper than in cortex , demonstrating that cortical auditory neurons do not directly inherit the already encoded spatial information present in lower centers of the auditory processing stream [1] , [28] . However , our results show that the spatial configuration of competing sources strongly affects the coding of those sources' content . Spatial effects in cortical neurons are far greater when there are competing sounds than when there is only a single source . This observation suggests that spatial information acts to modulate competition between sources , even in an anesthetized preparation . The fact that these effects arise under anesthesia is important because it shows that they are preattentive . Competition between spatially separated sources helps segregate neural responses , so that information about competing sound sources from different locations is concentrated in distinct subpopulations of cortical neurons . Specifically , most neurons preferentially encode information about contralateral sources; however , some neurons show more specific preferences . Thus , even though location is not directly coded in cortical neurons , spatial information strongly modulates cortical responses . This idea fits with recent results showing that the effects of source location on neurons in cortex are enhanced when an awake animal is engaged in a task requiring a localization response [10] . The degree to which spatial information affects cortical responses changes with intention: depending on the importance of spatial information to the task being undertaken , spatial coding may be either enhanced or weakened . It is possible that inhibition driven by activity in prefrontal cortex ( in mammals ) or its analog ( in avian species as studied here [29] ) causes the sharpened tuning observed during spatial tasks [10]; if so , such connections may also be engaged during selective attention tasks to down-regulate responses of neurons preferentially encoding a masking stimulus that is to be ignored or to up-regulate responses of the distinct population of neurons preferentially encoding the target . In an anesthetized preparation like that tested here , the enhancement of spatial effects due to the presence of a competing source cannot be coming from top-down modulation from executive centers of the brain . Instead , these effects must be the result of neural circuitry that is “hard wired . ” It may be that weak spatial tuning , which is not strong enough to cause observable changes in neural responses with changes in the location of a single sound source played alone , causes large effects when there are multiple sources from different positions . The preattentive spatial competition we find provides a substrate to realize selective spatial auditory attention . Once responses to competing sounds are partially segregated through this kind of preattentive , spatially sensitive process , attentional signals , including inhibitory feedback from executive control areas , can enhance the spatial selectivity already present . In humans , many spatial effects are explained by the fact that the head causes a significant acoustic shadow at many audible frequencies [26] , [30] . When competing sound sources come from different azimuthal locations , the SNR at the ear closer to the target will be greater than if the sources were co-located . This kind of “better-ear” effect has nothing to do with neural processing but is a simple consequence of physics . For the human , such effects can be very significant for speech perception , because the head shadow can be 15 dB or more for frequencies important for speech . Thus , although not interesting from a neural processing perspective , these acoustic effects are important for perception . Here , in the zebra finch , better-ear effects are small . The zebra finch head is diminutive; its width corresponds to only one quarter of the wavelength of the highest frequency present in our bandlimited stimuli ( 8 kHz ) . Given that appreciable acoustic interactions only arise when the wavelength of the sound is comparable to or smaller than the size of the physical object in the environment , the stimuli we presented did not contain frequencies high enough to cause large interaural level differences . This bears out in our measurements , which show an amplitude difference between the ears of approximately 1 . 5 dB when the stimulus is at ±90° . The better-ear effect is thus limited to 3 dB . Performance in the target contralateral , masker ipsilateral configuration was 16 . 8% better than performance in the target ipsilateral , masker contralateral configuration , on average . In contrast , the performance benefit of lowering the masker noise level by 6 dB is only 8 . 8% . Moving the masker and target in space , then , has nearly double the effect on identification performance as a 6 dB increase in SNR . Given that the maximum effect of acoustic head shadow is only 3 dB , the better-ear acoustic effects cannot explain the spatial effects obtained . Moreover , although a better-ear effect may contribute to the processing of natural broadband signals that contain frequencies high enough to interact acoustically with the zebra finch head , it is unlikely to play a major role in the effects observed here , where we used low-pass filtered stimuli . Interference from a masker on the response encoding a target can be broken down into two forms: spike additions ( primarily in the gaps between syllables ) and spike subtractions ( primarily during syllables ) [18] . Here , we quantified the effects of spatial configuration on spike additions and subtractions , and then evaluated modeled spike trains to determine the relative impacts of these effects on neural discrimination performance . In general , additions were more likely than subtractions when the target was ipsilateral to the recording site and masker was contralateral ( see Figure 4C ) , while subtractions were the more prevalent form of interference in the reverse configuration . Because additions and subtractions were calculated by comparing the responses at each spatial configuration to responses to the corresponding target-only stimulus , they represent only the effect of the masker on the response , independent of the minor changes that occur due to absolute target location . By modeling spike trains with only additions or only subtractions , we were able to gauge their effects on performance . Spike subtractions degraded performance at all configurations ( in Figure 4C , blue bars are lower than white bars ) . In contrast , the spike trains with only additive interference coded target song identity nearly as well as the responses in quiet ( red bars are nearly the same as white bars ) . Additions have a modest impact when subtractions are also present; additions-only performance was better than the fully masked responses in some configurations ( compare blue and black bars ) . Subtractions , on the other hand , interfere with encoding of song identity more seriously and consistently across all spatial configurations . Although this analysis does not reveal the mechanisms by which a masker interferes with coding of a target , it does give some insight into the complex interactions that take place when two competing sounds are present in an environment . For instance , one might expect , a priori , that the presence of an ongoing masker would cause activity to increase overall , so that the stereotypical target response in quiet is hidden amidst added spikes elicited by the masker . Yet , instead , the detrimental effects of the masker come about primarily from suppression of responses to key features in the target; moreover , the influence of the masker on the target response depends on spatial configuration . This pattern of spatial-configuration-dependent suppression of spikes suggests that competing sources , each preferentially encoded by a distinct neural subpopulation , mutually suppress each other , giving rise to enhanced spatial modulation of responses compared to when a single , unchallenged sound source is presented in isolation . For a given site , song identity coding tended to vary with both the target and masker locations and generally was best when the target was contralateral from the recording electrode and the masker was ipsilateral to it . For a single site to show spatial release from masking , performance for that site should increase monotonically with increasing spatial separation between the sources . Thus , neither any single recording site nor mean performance averaged over all sites ( shown in Figure 3B ) exhibits spatial release from masking . Similar results have been seen in the midbrain , in inferior colliculus [28] , where , as here , single units showed preferences for encoding responses to different sources , depending on the spatial configuration . However , the activity of thousands of forebrain neurons , not just a single unit , combines to govern perception and behavior . As shown in Figure 3 , across the population of neurons in forebrain , there are typically neurons contralateral to the target source that encode target identity well . By looking at the mean performance of neurons at recording sites for which the target sound is more contralateral than the masker ( or at the best neuron in that population ) , performance is predicted to improve with spatial separation ( see Figure 3C ) . Thus , the ensemble of responses , even from an anesthetized bird , can explain behavioral spatial unmasking if one assumes a mechanism as simple as attending to neurons in the hemisphere that favors encoding of the target and ignoring those from the opposite hemisphere . In behavioral experiments , performance improves with increasing separation between target and masker sounds both for speech and for non-speech sounds [26] , [31] , [32] . As noted above , better-ear acoustics contribute to spatial release from masking for many sounds important to human behavior , such as speech . Indeed , when a target sound is easily distinguished from a masker ( such as when a communication signal is played in steady-state noise ) , better-ear acoustics can fully account for spatial release from masking in human studies . Interestingly , avian studies do not show the same pattern . The amount of spatial release from masking is essentially identical when behaving birds identify target birdsongs embedded in either a chorus of songs that sound qualitatively like the target songs or a steady-state masker ( with the same long-term spectral content as the chorus , but has different short-term structure ) [27] . This different pattern suggests that humans can segregate a target from a dissimilar masker even when the two sources are near each other in space , rendering spatial cues redundant [33] . In contrast , birds may be less sophisticated in segregating competing sources , relying more heavily on spatial attributes even when target and masker have distinct spectro-temporal content . Regardless , the current results demonstrate how spatial separation of target and masker can support spatial release from masking in those situations where it is observed behaviorally , no matter what species .
All experimental procedures involving animals were done in accordance with the protocol approved by the Boston University Institutional Animal Care and Use Committee . All subjects were male zebra finches ( Taeniopygia guttata ) . Prior to the day of recording , a preparatory surgery was performed . In this surgery , the location of field L was marked as the point 1 . 2 mm anterior and 1 . 5 mm lateral of the midsagittal sinus and a headpin was fixed to the skull . On the day of recording , the bird was first placed in a soft cloth restraining jacket in a quiet , dark room . Injections of urethane anesthetic ( 20% ) were administered every half hour in decreasing amounts ( starting with 35 µL ) until the bird was unresponsive to its head being patted and its foot being squeezed . Once anesthetized , the bird was placed in a stereotactic frame with its head secured by the previously implanted pin . A craniotomy was performed in which an approximately 2 mm square of skull was removed centered about the spot previously marked as field L . Tungsten microelectrodes ( FHC , Bowdoin , ME ) ranging in impedance from 2 to 4 MΩ were advanced into the brain using a micron-precision stepper motor . Extracellular potentials were amplified at the headstage , bandpassed between 500 and 10 , 000 Hz , and recorded with a low-noise soundcard at a sampling rate of 44 . 1 kHz . Stimuli were constructed from combinations of two different target zebra finch songs and masking noise ( see Figure 1A and B for spectrograms ) , all filtered between 500 and 8 , 000 Hz . The songs were chosen to have similar durations ( ∼2 s ) ; they were songs never before heard by the subjects . To generate the masking noise , several songs were concatenated , the discrete Fourier transform computed , the phase randomized uniformly between 0 and 2π ( preserving symmetry ) , and the inverse Fourier transform computed . The result was noise with a magnitude spectrum identical to the average of the spectra of those songs , but with no temporal structure . Ten independent , random tokens of noise were created so that any residual temporal structure was averaged out across repeated presentations . Independent noise tokens were used on each trial instead of using a single , frozen token because individual noise tokens with the same statistics can have drastically different masking effects [34] . Additionally , the use of independent masker tokens better simulates what happens in natural settings , where , over time , a bird repeatedly hears highly stereotyped songs from its familiar colony mates , but hears them in a different background of masking sources each time . Stimuli were presented using four single-driver loudspeakers in a sound-treated booth ( IAC , Winchester , UK ) at a sampling rate of 44 . 1 kHz . Target songs were normalized so that their root-mean-square amplitudes were 72 dB SPL ( c-weighted ) . The loudspeakers were at four locations in the azimuthal plane: ipsilateral to the implanted hemisphere ( −90° ) , in front of the bird ( 0° ) , contralateral to the implant ( +90° ) , and at the angle halfway between the front and contralateral angles ( +45° ) . The speaker locations were referenced relative to the recording electrode , with the side ipsilateral to the implant assigned the negative sign . Each recording session consisted of 10 blocks . In each block , each of the two target songs was played in isolation from all four locations . Additionally , for each target song , 16 target-masker spatial configurations were tested , each at three SNRs . This resulted in 2× ( 4+4×4×3 ) = 104 stimuli per block in which targets were present . We also played the masker alone from each location in each block , resulting in a total of 108 stimuli per block . Each of the 10 blocks used a different , independent token of masking noise semi; the order of the stimuli within each block was randomized . Overall , there were 1 , 080 two-second stimuli presented with 1 . 5 s between the end of one trial and the beginning of the next , resulting in a recording session that lasted 63 min for each neural site . Extraction of action potentials ( spikes ) was performed off-line . First , neural traces were thresholded . The recording to 1 ms on either side of each local maximum was windowed out and considered a potential spike . These waveforms were sorted into user-defined template spike waveforms using a correlation-like coefficient:where xS is a spike waveform and xT a template waveform , and the sums are taken over time . Spikes were sorted into classes based on the template that yielded the highest r or were thrown out if they were not above a minimum r to any of the templates . This sorting was verified using principal components analysis clustering . Using this method , single units as well as multiunit clusters ( which could not be separated into single units ) were extracted . Of the sites that met the minimum performance criterion ( see below ) , 17 were single units and 16 were multiunit clusters . Multiunit activity should produce weaker spatial effects than well-isolated single units . By including both isolated and multiunit recordings in our analysis , our approach is likely to underestimate ( if anything ) the effects of spatial configuration on neurons in the forebrain . Recordings were made in both hemispheres , but a relative coordinate system was used so that negative azimuths always correspond to the hemisphere ipsilateral the recording site and positive azimuths to the contralateral side . Discrimination performance was calculated using a nearest-neighbor template-matching scheme and a spike distance metric . Methods used were similar to those used in past studies [17]–[19] , [22] , [24] , [35] , [36] . To compute pairwise distances between recorded spike trains , each spike train was convolved with a decaying exponential kernel whose time constant determined the effective integration time of the spike comparisons; then the sum of the squared difference was calculated . These distances were used to compare a test spike train against two templates , one from each target song . Each spike train was classified as being elicited by the song whose template was closest to the measured spike train . This process was repeated many times for all spatial configurations for kernel time constants of 1 , 4 , 16 , 63 , 251 , and 1 , 000 ms . All but one of the recording sites had an optimal time constant of 16 or 63 ms , in the same range as time constants found in similar past studies ( the outlier had an optimal time constant of 251 ms ) [17] , [22] . In this way , a percent correct score was calculated as a function of time constant , representing how well the spike trains from each spatial configuration matched the target spike trains from the template configuration . The time constant that yielded the best clean target discrimination for each site was used . Spatial sensitivity was computed as the difference between the maximum and minimum discrimination performance for a given stimulus type . For clean songs , these extrema were determined across the four target locations . For masked stimuli , they were determined across all 16 location configurations , at the SNR that had the highest variance . So that sites with poor performance did not appear spuriously insensitive to spatial configuration , only sites that had an unmasked discrimination performance of 90% or more were included in the analysis . To analyze the effects of spike additions and deletions separately , modeled spike trains were generated . Spike trains were binned into time bins of 2 . 5 ms . To generate a new spike train , the mean and standard deviation of the number of spikes in each time bin were computed across the 10 responses to each stimulus . Then , the number of spikes in each bin was chosen randomly from a Gaussian distribution with the same mean and standard deviation , with negative spike counts fixed to zero . Time bins in which the masked rate was higher than the clean rate were labeled “additions . ” Thus , to simulate a spike train that had only additions , the higher of the masked and clean rates ( and the corresponding SD ) was chosen at each time bin . Similarly , to simulate a subtractions-only spike train , the minimum of the masked and clean rates was chosen for each bin . The discrimination performance of these simulated spike trains was then calculated in the same manner as real recordings , described above . Refractory period violations ( interspike intervals of less than 1 ms ) had a negligible effect on analysis . The significance of the difference between the spatial sensitivity for clean and masked responses was computed using a paired Student's t test . All correlation r values were computed using Pearson's product-moment coefficient , with p values calculated using a Student's t distribution . The significance of spatial separation and SNR for the data shown in Figure 3C were computed using a two-way repeated measures ANOVA . After using a one-way ANOVA to confirm a significant effect of interference type ( p< . 001 ) , Tukey's HSD test was used to compute post hoc comparisons between all performance values at each of the three spatial configurations in Figure 4G . All statistics were done using Matlab's built-in functions . A p value of . 05 or less was considered significant .
|
When a listener is presented with many sound sources at once , it is easier to understand a particular source when it comes from a different spatial location than the other competing sources . However , past studies of auditory cortex generally find that in response to a single sound source , there is not a precise representation of spatial location in the cortex , which makes this effect of spatial location hard to understand . Here , we presented zebra finches with two simultaneous sounds ( a birdsong target and a noise masking sound ) from distinct spatial locations and recorded neural responses in field L , which is analogous to primary auditory cortex in mammals . When the target sound was presented by itself , the location of the source had little effect on the ability to identify the target song based on neural activity in field L . However , when the target was presented with a masker sound , the location of both sources strongly affected neural discrimination performance . Moreover , different subpopulations of neurons preferentially encoded either target or masker , providing a potential substrate for spatial selective attention . Thus , even though location is not well coded in cortical neurons , spatial information strongly modulates cortical responses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"auditory",
"system",
"biology",
"sensory",
"systems",
"sensory",
"perception",
"neuroscience"
] |
2012
|
Competing Sound Sources Reveal Spatial Effects in Cortical Processing
|
The interaction between intestinal parasites and malaria is still not clear . Data in published literature are conflicting . We studied the effect of intestinal parasitic infection ( IPI ) on the clinical outcome of malaria in coinfected children . In a cross sectional study performed between October 2014 and September 2015 , children infected with malaria , as demonstrated by the presence of asexual parasites in Giemsa stained blood films , were enrolled . Stool samples were obtained from participants and subjected to the formol-ether concentration technique for the detection of intestinal parasites . The Complete blood count was performed using an automated haematology analyser ( Mindray , BC-2800 ) . The risk ratio , Pearson’s chi-square and the student T test were all performed as part of the statistical analyses . Statistical significance was set at p < 0 . 05 . In all , 405 children successfully took part in the study . The children were between 1 week and 120 months of age ( mean ± SD = 41 . 5 ± 33 . 5 ) . Coinfection with intestinal parasites was observed in 11 . 6% . The rate of severe malaria ( SM ) attack in this study was 10 . 9% . SM was not observed to be associated with age ( p = 0 . 377 ) or gender ( p = 0 . 387 ) , meanwhile coinfection with intestinal parasites was associated with age ( p = 0 . 003 ) . Among SM cases , IPI prevalence was higher in children with mild ( WHO group 3 ) severe malaria ( p = 0 . 027 ) . Overall , IPI was not observed to be associated with SM ( p = 0 . 656 ) or malaria parasite density ( p = 0 . 185 ) or haemoglobin concentration ( p = 0 . 205 ) . The main clinical features of SM observed were hyperpyrexia ( 68 . 2% ) , severe malarial anaemia ( 61 . 4% ) , and multiple convulsion ( 52 . 3% ) . IPI was not observed to be associated with the severity of malaria , the malaria parasite density , and the haemoglobin concentration in coinfected children in Cameroon . The clinical outcome of malaria in children coinfected with intestinal parasites may depend on the geographical setting after all .
Malaria and intestinal parasitic infections ( IPIs ) are parasitic diseases that are highly endemic in Sub-Saharan Africa , especially in impoverished and poor sanitary settings . In 2013 , there were 198 million cases and 584000 deaths as a result of malaria worldwide [1] . The majority of deaths due to malaria are reported in Sub-Saharan Africa and occur mostly in children below 15years [2] . Five species of Plasmodium are known to cause malaria in humans including P . falciparum , P . ovale , P . vivax , P . malariae , and P . knowlesi . Among them , P . falciparum is the most virulent species responsible for a majority of cases and almost all malaria associated deaths [3 , 4] . Like malaria , the prevalence of IPIs is also higher in developing countries , reaching up to 95% in some settings in Sub-Saharan Africa . IPIs are caused by either helminths , protozoa or both . Because of the overlapping distribution of malaria and IPIs , coinfection between malaria and IPI are therefore common in Sub-Saharan Africa [5] . The interaction between malaria and the different intestinal parasites whenever coinfection is present is poorly understood . Infection with intestinal parasites especially helminths is strongly suspected to influence the incidence , parasite density and the clinical outcome of malaria in endemic areas . Earlier studies have reported increased susceptibility to malaria [6–8] , increased malaria gametocyte carriage [9] , decreased haemoglobin concentration [10] and increased risk for clinical and severe malaria [11–13] in helminth coinfected individuals . Conversely , other studies have reported that helminth infections may protect from malaria , or related clinical outcomes by suppressing acute clinical manifestations [14 , 15] , parasite density [16 , 17] or severe complication such as cerebral malaria [18] , circulatory collapse [19] , renal failure and jaundice [20] . Yet in other studies , no significant association have been observed [7 , 21] . The variations in the results obtained could be attributed to the different helminth species , different transmission settings , and the complex nature of the immune responses to malaria parasites or the altered immune response due to helminth co-infections . In addition , variations in the study design or methodology , case definition or malaria severity status , stage or intensity of species of helminths or Plasmodium and other confounding factors could equally contribute to the varying results observed in these studies [22–24] . In Cameroon , like most Sub-Saharan African countries , coinfection between malaria and intestinal parasites is common in children [25–28] . Across the country , the prevalence of coinfection with malaria and intestinal parasites varies from 11 . 9% to 34 . 7% [25–29] and school-age children are the most affected . IPI may exacerbate the severity of malaria in these children . To the best of our knowledge , no study has been performed in Cameroon to determine the interaction between malaria and intestinal parasitic infection . This study was therefore designed to give an insight on the effect of IPIs on the clinical outcome of malaria ( such as severe malaria and parasite density ) in coinfected children in Cameroon .
This was a cross sectional study performed between October 2014 and September 2015 . Buea ( 4°10′0″N 9°14′0″E ) with an elevation of 870m ( 2 , 850ft ) is located in the eastern slopes of Mount Cameroon . Buea is the capital of the Southwest Region of Cameroon . The population of Buea is estimated at 200 , 000 [30] and it is considered one of the fastest growing towns in Cameroon today with a mix cosmopolitan setting and a constellation of about 67 villages . The major ethnic group is the Bakweri ( the indigenes ) . Because of its location at the foot of Mount Cameroon , the climate in Buea tends to be humid , with the neighbourhoods at higher elevations enjoying cooler temperatures while the lower neighbourhoods experience a hotter climate . Extended periods of rainfall , characterized by incessant drizzle , which can last for weeks , are common during the rainy season . The planning of Buea especially the villages is poorly done and several breeding sites for Anopheles mosquitoes can be seen around homes . Buea has two seasons; the dry season ( between October and March ) , and the rainy season ( between April and September ) . In Buea , human malaria can be described as mesoendemic in the dry season and hyperendemic in the rainy season , with peaks at the beginning and towards the end of the rainy season . The population of Buea experiences an estimated 3 . 93 infective bites/person/night [31] . Plasmodium falciparum accounts for up to 96% of malaria infections in this area [32] . In addition , Buea is also highly endemic for intestinal parasites including Ascaris lumbricoides , hookworm etc . [33 , 34] . School-age children are the most affected and as a result , mass deworming campaigns are organized annually [35] to curb down the burden . Febrile children ( ≤10 years ) who came to consult in the outpatient department or emergency unit of the Regional Hospital of Buea were considered . The vital signs of the patients were taken and they were examined by the consulting physicians . Patients that were positive for malaria parasites by light microscopy were enrolled . Excluded from the study were patients with a history of anti-malarial treatment within one week prior to admission . Patients were also excluded from the study based on evidence of other infectious disease , such as typhoid , gastroenteritis , meningitis , malnutrition , upper respiratory tract infections or any other identified cause of anaemia other than malaria . The study protocol was approved by the Institutional Review Board ( IRB ) of the Faculty of Health Sciences , University of Buea , Cameroon . Administrative clearance was obtained from the delegation of public health in the South West region of Cameroon . Written informed consent was obtained from the parents or guardians of the patients prior to their inclusion into the study . Blood and stool specimen were collected from the participants . About 3ml of blood specimen was collected into test tubes containing EDTA anticoagulant following aseptic techniques . The parents or guardians were instructed to provide a teaspoon full of their child’s stool into a sterile leak-proof wide open neck stool container . Thick and thin blood films were prepared and stained with 10% Giemsa and examined using methods previously described [36] . If parasites were observed , the density was determined by counting the number of parasites against 500 leucocytes . The parasite density was obtained by dividing the number of parasites counted by 500 and multiplying the result by the actual WBC count of the patient [37] . The haemoglobin concentration ( Hb ) was obtained from the complete blood count ( CBC ) results of the patient . The CBC was performed using the Mindray Auto haematology analyzer ( BC-2800 , Shenzhen Mindray Bio-Medical Electronics Co . , Ltd ) . The formol ether concentration technique ( FECT ) was used for the detection and quantification of intestinal parasites in stool samples from participants . Since hookworm eggs clear very rapidly , the FECT was performed within 30mins upon receipt of stool specimen following the proceedings described by Cheesbrough [38] . Briefly , about 1g of stool from each sample was emulsified in 4ml of 10% formalin in a conical tube with the aid of an applicator stick , to obtain a slightly cloudy suspension . A further 3ml of 10% formalin was added and thoroughly mixed by shaking ( for approx . 30secs ) . The content of the tube was then strained through a surgical gauze into a beaker . The suspension was then transferred into another conical tube and 3ml of diethyl ether was added to it . The tube was closed with a rubber stopper and shaken vigorously for 1minute before centrifuging at 3000rpm for 5minutes . After spinning , the fatty plug ( debris ) was released from the sides of the tube with the aid of an applicator stick , and the supernatant was then poured away by quickly inverting the tube . The resulting sediment was then mixed by gently tapping the tubes with the fingers and the sediment was then transferred to a glass slide and covered with a cover glass . The whole area under the cover glass was then examined for ova , cysts , and larvae using the 10x and 40x objective of a light microscope . To assist in the identification of cysts , a small drop of iodine was added to the preparation under the cover glass . Helminths ova were separately counted for each species and recorded as previously described [39] . Helminths ova were counted until 100 ova were reached . Counts exceeding 100 were recorded as ‘100+’ . A semi quantitative scheme was adopted for intestinal protozoa whereby samples were recorded as ( i ) ‘negative’ in cases where no cyst or trophozoite was observed in the entire sediment; ( ii ) ‘rare’ in cases where one to five cysts or trophozoites were observed per slide; ( iii ) ‘frequent’ in cases where one cyst or trophozoite was observed per field at 400x magnification; and ( iv ) ‘very frequent’ in cases where over one cysts or trophozoites were observed per field at 400x magnification . Severe malaria ( SM ) was defined and categorized base on the WHO [40] criteria as follows; ( 1 ) severe anaemia ( Hb<5g/dl ) with no history of severe bleeding; ( 2 ) prostration , defined as the inability to sit upright or eat for a child who was able to do so; ( 3 ) respiratory distress , defined as difficulty in breathing with characteristic nasal flaring , subcostal recessions; ( 4 ) multiple convulsion reported within the preceding 24hrs plus one directly observed; ( 5 ) impaired consciousness , defined as a score ≤4 on the Blantyre Coma Scale [41]; ( 6 ) clinical jaundice; ( 7 ) circulatory collapse , defined as a systolic blood pressure <60 mm of Hg in children ≤5 years or < 80 mm of Hg in children > 5 years , in addition to observation of weak or absent of peripheral pulses or cold limbs; ( 8 ) abnormal bleeding; ( 9 ) pulmonary edema; and ( 10 ) frequent vomiting [42] . The term cerebral malaria was reserved for Blantyre coma score ≤2 corrected for no record of recent severe head trauma , neurological disease or any other cause of coma . And uncomplicated malaria ( UM ) defined by being fully conscious with haemoglobin ≥ 8g/dl and no signs of severity and/or evidence of vital organ dysfunction . SM was further classified according to the degree of severity using the WHO scheme ( S1 Table ) into 3 groups with the first group being the most severe and associated with the highest mortality rate , the second group being moderately severe and the third being the least severe . Data collected were keyed into Excel spreadsheets and analysed using Stata version 12 . 1 ( StataCorp LP ) . Statistical tests performed included the risk ratio , and the Pearson’s Chi square test for group comparison , student T test for the comparison of group means . Statistical significance was set at p < 0 . 05 .
Out of the 637 children who were approached , 405 children met the inclusion criteria and were therefore enrolled . The ages of the children ranged between 1 week and 120 months ( mean±SD = 41 . 5±33 . 5 ) . Among the participants were 209 ( 51 . 6% ) females and 196 ( 48 . 4% ) males . Among the participants , 398 ( 98 . 3% ) had single infection with Plasmodium falciparum , 5 ( 1 . 2% ) had mixed infection with P . falciparum + P . malariae , and 2 ( 0 . 5% ) had mixed infection with P . falciparum + P . ovale . Severe malaria ( SM ) was observed in 44 ( 10 . 9% ) of the 405 participants . All ( 100% ) SM cases had single infection with Plasmodium falciparum . No significant association was observed between severe malaria and age ( p = 0 . 377 ) nor was there any significant association between SM and gender ( p = 0 . 387 ) ( Table 1 ) . Stratified according to the degree of severity of SM , the distribution of the cases in decreasing order were: 26 ( 55 . 3% ) for group 1; 17 ( 36 . 2% ) for group 2; and 4 ( 8 . 5% ) for group 3 . The main clinical presentation of SM in decreasing frequency were hyperpyrexia 30 ( 68 . 2% ) , severe malarial anaemia 27 ( 61 . 4% ) , multiple convulsion 23 ( 52 . 3% ) , circulatory collapse 19 ( 43 . 2% ) , respiratory distress 16 ( 36 . 4% ) , jaundice 15 ( 34 . 1% ) , protraction 11 ( 25% ) , hyperparasitaemia 7 ( 15 . 9% ) , impaired consciousness 7 ( 15 . 9% ) , hypoglycaemia 6 ( 13 . 6% ) , cerebral malaria ( BCS≤2 ) 5 ( 11 . 4% ) , frequent vomiting 4 ( 9 . 1% ) , and coma 2 ( 4 . 6% ) . The overlap of severe malaria anaemia with cerebral malaria and respiratory distress were observed in 4 . 6% ( 2/44 ) and 15 . 9% ( 7/44 ) participants respectively ( Fig 1 ) . One of the 44 ( 2 . 3% ) participants with SM presented with all the three major clinical subgroups of malaria ( severe malarial anaemia , cerebral malaria and respiratory distress ) . Among the 405 participants , 47 were coinfected with intestinal parasites giving a prevalence of 11 . 6% ( 95% CI: 8 . 7–15 . 1 ) . No significant association was observed between the prevalence of coinfection and gender ( p = 0 . 484 ) ( Table 2 ) . On the contrary , there was a significant association between prevalence of coinfection and age ( p = 0 . 003 ) ( Table 2 ) . The intestinal parasite isolated in decreasing order of prevalence were Ascaris lumbricoides 36 ( 69 . 2% ) , Entamoeba spp . 12 ( 23 . 1% ) , and Hookworm 4 ( 7 . 7% ) . Among the coinfected children , 6/47 ( 12 . 8% ) had SM meanwhile among non coinfected children , 38/358 had SM ( 10 . 6% ) . However no significant association was observed between the presence of coinfection and SM ( χ² = 0 . 20 , p = 0 . 656 ) ( Table 3 ) . Stratified according to the degree of severity , the prevalence of coinfection was highest among cases of group 3 ( mild SM ) of the WHO classification ( Table 4 ) . A significant association was observed between the degree of severity of SM and the prevalence of coinfection ( χ² = 7 . 21 , p = 0 . 027 ) . Species-specific analysis revealed no significant association between the different species of intestinal parasites and SM ( χ² = 4 . 99 , p = 0 . 172 ) . Furthermore , the risk of developing SM was similar between all intestinal parasite species ( Table 5 ) . The geometric mean parasite density ( GMPD ) in this study was 10332 . 7 ( range: 65–160523 ) . The GMPD was higher in children without IPI 10732 . 1 ( range: 65–160523 ) compared to coinfected children 7290 . 6 ( range: 171–82956 ) . However this difference was not observed to be significant statistically ( p = 0 . 185 ) . The average intensity ( epg ) of intestinal helminth was 28 . 2 ( range: 2–134 ) . The average intensity of intestinal helminths was higher in SM cases 36 . 8 ( range: 8–105 ) compared to UM cases 26 . 9 ( range: 2–134 ) ( Table 6 ) . However this difference was not observed to be significant ( p = 0 . 272 ) . The intensity of infection with Entamoeba spp . was generally rare across all groups ( Table 6 ) . The mean Hb observed in this study was 10 . 6 ( ±1 . 8 ) . The mean Hb was lower in children with IPI ( 10 . 4g/dl ) compared to children without IPI ( 10 . 7g/dl ) ( Fig 2 ) . However the difference in the mean Hb between children with IPI and those without was not observed to be significant statistically ( p = 0 . 205 ) . The prevalence of anaemia were 44 . 4% ( 16/36 ) , 50 . 0% ( 2/4 ) , 57 . 1% ( 4/7 ) and 60 . 0% ( 3/5 ) in participants infected with A . lumbricoides , hookworm , Entamoeba spp . and mixed A . lumbricoides/Entamoeba spp . respectively . However no significant association was observed between anaemia and the infecting parasite species ( χ2 = 0 . 18 , p = 0 . 980 ) .
In this study , 11 . 6% of the participants were coinfected with intestinal parasites . These findings are similar to that reported by Njunda et al . [28] . The prevalence is however lower compared to the 34 . 7% reported in communities around Dschang in the West region of Cameroon [43] . The difference in the prevalence reported in Dschang and this study could be attributed to differences in the study design; our study targeted children in an urban setting meanwhile theirs targeted mainly school-age children residing in rural settings . The prevalence of intestinal parasites observed in this study is lower compared to the 22 . 7% reported in Thailand [44] , and the 34 . 2% reported in Ethiopia [45] . The lower prevalence of IPI observed in this study could be attributed to the regular deworming campaigns organized by the Cameroon’s Ministry of Public Health targeting mainly children in the study area . The prevalence of IPI was observed to be significantly higher in children in the age range 60–120 months , which is in line with studies performed elsewhere [28 , 43 , 46] . The higher prevalence of IPI in this age group could be attributed to the differences in the exposure as the children grow up , become more playful especially with soil . The most frequent intestinal parasite isolated was Ascaris lumbricoides ( 69 . 2% ) . Other studies have also shown that Ascaris lumbricoides is the most prevalent species of intestinal parasites isolated from children [25 , 46–49] . Hookworm infection is present in the study area but the prevalence observed is low ( 8 . 5% ) . The concentration method used in this study may have impacted the result of hookworm infection , therefore the real prevalence might have been underestimated . In the current study , it was observed that 10 . 86% of the participants had severe malaria ( SM ) based on the case definition of severe malaria by the WHO [40] . The prevalence of SM in this study is similar to the prevalence reported by Tchokoleu et al . [50] , but lower compared to the prevalence reported elsewhere in Cameroon [51 , 52] , and in other countries [53–56] . The differences in the prevalence of SM observed in these studies and ours could be due to the ages of the target population—our study involved children 10 years and below meanwhile most of the studies performed elsewhere focused on children 5 years and below . The risk of SM has been shown to be higher in children below 5 years compared to older children due to the low immunity , which has been shown to increase with increasing age [40] . Furthermore , the level of endemicity of malaria could be another factor responsible for the differences in the prevalence observed in our study and the others; the endemicity of human malaria in the study area varies between mesoendemic in the dry season and hyperendemic in the rainy season . Although the prevalence of SM was observed to be higher in children below 5 years and in males in this study , no significant association was observed between SM and age , neither was there any association between SM and gender . The observation of higher prevalence of SM in males corroborate the studies performed elsewhere [55 , 56] . In this study , P . falciparum was the only parasite species observed to be associated with SM . P . falciparum is well recognised as the main cause of SM in endemic areas . Although there are reports of P . vivax also causing SM [57] , P . vivax was not identified as a species causing malaria in this study . The main clinical features of SM in this study were hyperpyrexia , severe malaria anaemia and convulsion , which is similar to studies performed elsewhere [52 , 56] . In the current study , no significant difference was observed in the prevalence of severe malaria in children coinfected with intestinal parasites ( 12 . 8% ) compared to malaria monoinfected children ( 10 . 6% ) . This is contrary to studies associating IPI and protection from severe malaria [18 , 20] , as well as studies associating IPI and increased risk of SM [11–13] . Species-specific analysis also did not yield any significant association between the different intestinal parasite species and severe malaria ( p = 0 . 281 ) , which contradicts the study by Le Hesran et al . [12] in which infection with Ascaris lumbricoides was observed to be associated with increased attack of SM . This could be attributed to the differences in study design—our study was a cross sectional study involving children infected with malaria whereas theirs was a case-control study . Furthermore , no significant association was observed between the intensity of helminthic infection and SM in this study ( p = 0 . 272 ) . In this study , coinfection with malaria and intestinal parasites was significantly higher in group 3 ( i . e . mild SM ) of the WHO classification of SM ( p = 0 . 027 ) . The main phenotype of SM in this group is persistent vomiting; infection with malaria parasites or intestinal parasites are associated with gastrointestinal disturbances which may results in nausea and vomiting . Concomitant infection with both parasites may increase the progression of the gastrointestinal disturbances leading to increase in the frequency of vomiting . This observation should however be interpreted with caution as just a few ( 6 ) cases of IPI was observed among participants with SM . Larger studies will therefore be required to confirmed this association . In the present study , the geometric mean parasite density was higher in malaria monoinfected children ( 10732 . 1 parasites/ul ) compared to children coinfected with intestinal parasites ( 7290 . 6 parasites/ul ) . The interaction between intestinal parasites and malaria is not well understood . IPI especially helminths have been shown to have an immunomodulatory effect by inducing a Th2 host response . The host response will depend on the stage and intensity of helminth infection [57 , 58] . Clearly , the host response may also depend on the specific type of helminthic infection and the age of the host . However in this study , no significant difference was observed in the mean parasite density between children with IPI and those without IPI , which contradicts the study by Brutus et al . [16] in which treatment of Ascaris lumbricoides was observed to be associated with a 2-fold increase in malaria parasitaemia ( suggesting a protective effect ) . The failure to observe any significant difference in the mean malaria parasite density between children with IPI and those without could be attributed to the study design . The cross sectional design of the present study , cannot differentiate between present and past infection; the Th2 polarization of immune responses to intestinal helminths due to previous exposure may still be maintained through certain period although the infection itself has been cleared at present . The mean Hb observed in this study was low in children coinfected with IPI compared to children without IPI . IPI , especially infection with hookworm and Trichuris trichiura , cause anaemia by increasing blood and iron loss in the intestinal tract . However the difference in the mean Hb between IPI coinfected children and children without IPI was not observed to be significant ( p = 0 . 205 ) , which is contrary to the study by Nacher et al . [10] . Moreover , species specific analysis revealed no significant association between the infecting intestinal parasite species and anaemia . The failure to observe any significant difference in the mean Hb between children coinfected with IPI and those without could be attributed to the overall high prevalence of anaemia in this study as illustrated by the mean Hb 10 . 6 ( ±1 . 8 ) for which there are many other causes including malnutrition , which is very rampant in and around communities in Buea . The finding of no significant association between IPI and the clinical outcome of malaria in this study could be attributed to the low prevalence of IPI in the study area in addition to the other factors highlighted above . Studies involving larger number of cases with IPI are therefore required . In conclusion , 11 . 6% of the participants were coinfected with intestinal parasites . The rate of severe malaria attack observed in the study was 10 . 9% . Severe malaria was not observed to be associated with age or gender . The prevalence of IPI was observed to be significantly higher in children with mild severe malaria . The main clinical presentations of severe malaria were hyperpyrexia , severe malarial anaemia and multiple convulsion . IPI was not observed to influence the severity of malaria , the malaria parasite density and the haemoglobin concentration in children in the study area , which is contrary to many studies published in the scientific literature . As it may appear , the clinical outcome of malaria in children coinfected with intestinal parasites may depend on the geographical setting .
|
Coinfection with malaria and intestinal parasites are common in Sub-Saharan Africa , particularly in impoverished and poor sanitary settings . The interaction between intestinal parasites and malaria in coinfected children is still not clear . Some published papers suggest intestinal parasites , especially Ascaris lumbricoides , may attenuate the severity of malaria in the presence of coinfection . In this cross-sectional study , we evaluated the effect of intestinal parasitic infection on the severity of malaria , malaria parasite density and the haemoglobin concentration in children coinfected with malaria and intestinal parasites in Cameroon . We did not observe any significant association between intestinal parasitic infection and severe malaria or malaria parasite density or haemoglobin concentration . Stratification of severe malaria according to the degree of severity revealed a significant association with intestinal parasitic infection , in which prevalence of intestinal parasites was higher in children with mild severe malaria . Analyzing the different species of intestinal parasite did not yield any significant association either . These findings are contrary to many research publication on the subject . Several factors could have contributed to our observation , including the regular deworming campaign organized by the Cameroon Ministry of Public health , accounting for the lower prevalence of intestinal parasitic infection , and also the geographical setting .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"tropical",
"diseases",
"microbiology",
"geographical",
"locations",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"parasitology",
"protozoans",
"africa",
"cameroon",
"parasitic",
"intestinal",
"diseases",
"malarial",
"parasites",
"people",
"and",
"places",
"helminth",
"infections",
"virology",
"co-infections",
"biology",
"and",
"life",
"sciences",
"malaria",
"intestinal",
"parasites",
"organisms"
] |
2016
|
The Effect of Intestinal Parasitic Infection on the Clinical Outcome of Malaria in Coinfected Children in Cameroon
|
Adults combine information from different sensory modalities to estimate object properties such as size or location . This process is optimal in that ( i ) sensory information is weighted according to relative reliability: more reliable estimates have more influence on the combined estimate and ( ii ) the combined estimate is more reliable than the component uni-modal estimates . Previous studies suggest that optimal sensory integration does not emerge until around 10 years of age . Younger children rely on a single modality or combine information using inappropriate sensory weights . Children aged 4–11 and adults completed a simple audio-visual task in which they reported either the number of beeps or the number of flashes in uni-modal and bi-modal conditions . In bi-modal trials , beeps and flashes differed in number by 0 , 1 or 2 . Mutual interactions between the sensory signals were evident at all ages: the reported number of flashes was influenced by the number of simultaneously presented beeps and vice versa . Furthermore , for all ages , the relative strength of these interactions was predicted by the relative reliabilities of the two modalities , in other words , all observers weighted the signals appropriately . The degree of cross-modal interaction decreased with age: the youngest observers could not ignore the task-irrelevant modality—they fully combined vision and audition such that they perceived equal numbers of flashes and beeps for bi-modal stimuli . Older observers showed much smaller effects of the task-irrelevant modality . Do these interactions reflect optimal integration ? Full or partial cross-modal integration predicts improved reliability in bi-modal conditions . In contrast , switching between modalities reduces reliability . Model comparison suggests that older observers employed partial integration , whereas younger observers ( up to around 8 years ) did not integrate , but followed a sub-optimal switching strategy , responding according to either visual or auditory information on each trial .
Imagine you are at an academic conference . A heated debate turns nasty and one scientist is repeatedly hit before falling to the floor . You are later asked how many punches were thrown . You confidently answer ‘3’; you were able to combine information from audition and vision , having both seen and heard the incident . We often receive information about the same object or event from multiple sensory modalities that we can integrate to improve the precision of our perceptual estimates . As adults , we integrate multisensory information for a variety of spatial and temporal tasks , such as judging the size , location , number or duration of objects or events [1–5] . A key benefit of this integration is that uncertainty , or variance ( random noise ) in the combined , multisensory estimate is reduced , relative to either of the component uni-sensory estimates , see e . g . [6] . Under standard models of integration , sensory estimates are combined via weighted averaging , according to the estimates’ relative reliabilities , see e . g . [1 , 2 , 6] . For example , consider the case in which an aspect of the environment is estimated from vision and audition . The visual and auditory estimates , S^V and S^A are not perfectly precise , but contain noise with variance σV2 and σA2 . It is commonly assumed that these noise distributions are Gaussian and independent . Under these assumptions , and given that the prior probability distribution over the estimated variable is uniform , then the optimal audio-visual estimate ( i . e . that with the lowest possible variance ) , for a continuous variable is given by: S^VA=wVS^V+wAS^A with the visual and auditory weights , wV and wA defined as: wV=1/σV21/σV2+1/σA2 and wA=1/σA21/σV2+1/σA2 . As can be seen from the equations above , sensory weights give the relative influence of each uni-modal sensory estimate in determining responses to bi-modal stimuli . These weights can be estimated from behavioural data corresponding to bi-modal and uni-modal stimulus conditions . For example , in a size estimation task such as [1] , subjects might be required to estimate an object’s size from vision alone , from haptics ( touch ) alone , or from both vision and hatpics . If the visual size is 9cm and the haptic size is 12cm , then given unbiased uni-modal estimates , a mean bi-modal response of 10cm would correspond to visual and haptic weights of 2/3 and 1/3 , respectively , i . e . vision has double the influence of haptics . These observed weights would be optimal if the uni-modal visual estimates were twice as reliable as the uni-modal haptic estimates , i . e . σV2/σH2=0 . 5 . Observing optimal sensory weights is consistent with optimal integration , i . e . the integration behaviour that minimises variance in the multimodal estimates . However , optimal sensory weights might be observed in the absence of integration: as an alternative to integration , an observer may select one of the uni-modal estimates on each trial , rather than computing a weighted average [7 , 8] . In the example above , the observer may select the visual estimate on 2/3 of trials , and the haptic estimate on 1/3 of trials . This ‘switching’ behaviour would produce the same mean response in bi-modal conditions as optimal integration , but with higher variance . Standard models predict that variance will be reduced in bi-modal , relative to uni-modal conditions under optimal integration , see , e . g . [1 , 2 , 6 , 9] . For example in the visual-haptic size example , under optimal integration the predicted variance of the visual-haptic estimates , σVH2 , is given by σVH2=σV2σH2σV2+σH2 . In contrast , switching behaviour will result in variance that is at least as large as the more reliable cue . For this reason , studies of multimodal integration generally determine ( i ) whether the sensory weights are optimal , given uni-sensory variability , and ( ii ) whether variability in the bi-modal estimates is reduced , relative to uni-modal estimates . Recently , a number of studies have asked whether children show optimal integration of sensory cues , as indexed by ( i ) appropriate cue weighting and ( ii ) a reduction in variance , relative to single cue conditions . Gori and Burr [10] reported that optimal integration of multisensory information doesn’t appear until surprisingly late—at the age of around 10 years . In two visual-haptic tasks , younger children who were asked to judge object size or orientation relied on only one modality , and not necessarily the most reliable one . Other work has confirmed that children as old as 8 years fail to optimally integrate visual cues with movement-based information ( proprioceptive and vestibular ) for navigation [7] , and another study suggests that optimal integration of auditory and haptic information does not occur until after age 11 [11] . Interestingly , this developmentally late integration is not limited to situations in which information must be combined from different sensory modalities: Nardini and colleagues reported similarly late integration for cues within a modality—optimal integration of two visual depth cues did not emerge until around age 12 [12] . The current study focuses on the developmental trajectory of audio-visual integration , using a straightforward counting task . The age at which optimal integration emerges for vision and audition is not yet clear . One previous audio-visual study with children aged 5–14 years and adults failed to find optimal integration at any age [13] . We employed a simple audio-visual task in which , on each trial , observers were presented with a number of beeps and / or flashes [14] . In separate blocks , they either reported the number of flashes , or the number of beeps . The task had the benefit of reduced memory and decisional demands , relative to previous studies that have used two-alternative forced choice designs . By comparing observers’ responses to different integration models we ask:
Three models were compared: ( i ) Partial Integration , ( ii ) Focal Switching , and ( iii ) Modality Switching . Note that these were evaluated separately for each observer; averaged fits are shown in Figs 4–7 for illustration only . For all three models , because the number of events can take integer values only , noise distributions , and the resultant uni-sensory likelihoods were approximated by discretised Gaussians , i . e . the probability of a sensory estimate equal to x , is given by p ( x ) =aexp ( − ( x−μ ) 22σ2 ) , {x ∈ℤ | ≥ 0} where a is a normalising constant . Noise distributions were centred on the true stimulus value , μ , but differed in variance , σ2 , for vision and audition ( see Fig 4 ) . In addition , alternative models were evaluated including the Causal Inference model [18 , 19] , models with logarithmic coding of number ( corresponding to skewed likelihoods in linear space ) , and those that allowed likelihoods to be biased and / or to vary in reliability as a function of the number of events ( beeps or flashes ) . These other models provided an inferior account of the data , as described in the supporting information file: S1 Text . The three models ( Partial Integration , Focal Switching , Modality Switching ) differ in the way that sensory information from vision and audition interact: For each observer and each model , the values of the 5 free parameters were found ( Matlab: fminsearch ) that maximised the joint likelihood of the observer’s data across all uni-modal and bi-modal conditions . To avoid the problem of local minima , 288 iterations of the search were performed , making use of the University of Southampton’s IRIDIS High Performance Computing facility , with initial values uniformly sampled from the multidimensional space of plausible parameters . Fig 8 shows how multisensory interactions change as a function of age . Observers in the two youngest groups were best described by the switching models . Children aged 8–9 years were evenly split , and by 10 years the majority of observers followed the partial integration model . As the three different models have common parameters , ( visual and auditory noise , and the mean and variance of the prior over the number of events ) we can consider how the best fitting values of these change as a function of age . Recent work [21] suggests that children as young as 7 quickly learn the statistics of a stimulus set and bias their estimates towards the mean . In the current study , knowledge of the stimulus statistics would be represented within the prior over the number of events . As participants learn these statistics we might expect both the mean and standard deviation of the prior to decrease , as participants learn that only a small numbers of beeps and / or flashes are presented . The youngest group had the weakest prior ( largest standard deviation ) of all age groups; this parameter varied significantly as a function of age ( F4 , 71 = 3 . 03 , p<0 . 05 ) . Post hoc comparisons showed that the youngest group had a significantly weaker prior than the 6–7 and 8–9 year olds ( p<0 . 05 from independent t-tests , after correction for multiple comparisons ) , no other comparisons were significant . Whilst the fitted prior for youngest group also had the largest mean , this did not vary significantly across groups . This provides some evidence that the youngest group may have been slower to learn the stimulus statistics . As might be expected from the raw response variance shown in Fig 3 , the fitted visual and auditory noise parameters also varied as a function of age ( F4 , 71 = 6 . 8 , p<0 . 001; F4 , 71 = 6 . 9 , p<0 . 0001 , for σV and σA , respectively ) . Visual noise decreased monotonically with age , auditory noise decreased across each age group pair that shared a common stimulus ISI . Posthoc t-tests showed that , based on the fitted noise parameters , the youngest group was significantly more variable in both vision and audition than all other groups p<0 . 01 , after corrections for multiple comparisons ) . In the current paradigm ISI decreased with age ( in order to broadly equate task difficulty ) . With a fixed ISI we would expect a larger increase in visual and auditory temporal acuity as a function of age .
A simple task was used to investigate the developmental trajectory of audio-visual integration . Importantly , we evaluated three different models of integration that together provide a good account of sensory integration behaviour at all stages of development . Key findings emerged: Sensory integration has the potential to provide benefits for virtually all of our everyday activities—precision is improved by combining redundant information sources either within or across modalities . An obvious question remains unanswered—why does this ability fail to appear until around 10 years ? One proposed explanation is that the lack of integration is beneficial during early childhood , and facilitates recalibration [10 , 12] . During this period of growth and sensory development , constant sensory recalibration is required in order to maintain accurate ( unbiased ) perceptual estimates . Recalibration requires the estimation of inter-sensory conflict—if this were only possible in the absence of integration , i . e . by keeping sensory estimates separate , then the developing sensory system might forego integration in favour of recalibration . The importance of cross-sensory interaction for sensory calibration and development is supported by studies in populations with sensory impairments—congenital visual deficits appear to have a detrimental effect on the precision of haptic estimates and vice versa [25 , 26] . Studies with adult observers , however , suggest that integration and recalibration are not mutually exclusive . For example , when glasses distort the relationship between binocular disparity and depth , the perceptual system recalibrates accordingly , whilst continuing to integrate binocular disparity with other depth cues [27] . Moreover , the sensory system adapts relatively quickly ( within hours ) when sensory statistics change [17 , 28–30] . In fact , recalibration and integration both rely on establishing the correspondence between signals—identifying which signals are redundant and only integrating ( or recalibrating ) when they arise from the same source . It might be that younger children find this correspondence hard to learn [31] . In the current study , observers were told to ignore one modality—adults were able to do this to a large extent , whereas children were sub-optimal in the sense that cross-modal influences were larger , even though vision and audition were discrepant on the majority of trials . A previous study also found that the effect of auditory beeps on the reported number of flashes was larger in children than adults [32] . However , that study did not use a design that allowed optimal integration to be evaluated . One recent study using a visual-proprioceptive reaching task did find some evidence of optimal integration , as evidenced by a reliability benefit , in children as young as 4–6 years [33] . However , this was only for the subset of observers who showed similar reliability for visual and proprioceptive estimates . Sub-optimal behaviour in other observers was attributed to inappropriate weighting . However , because the study did not include cue-conflict conditions , precise estimation of cue weightings was not possible . Our data suggest that , at least for the current task , the lack of integration shown by our observers was not due to a failure to weight the available signals appropriately . In summary , the current work suggests that optimal integration does not emerge until around 10 years . Model comparison suggests that before that age , observers switch between the information provided by the two modalities , but do so in accordance with their relative reliabilities . This behaviour does result in responses centred on optimal values , but variance is larger than under optimal integration . In contrast with previous work , our younger observers did not rely on a single modality—in fact they were less able to ignore task-irrelevant information . Instead , they instead showed stronger , mandatory cross-sensory interactions than older observers .
Visual stimuli were white discs subtending 2 . 2 degrees of visual angle ( dva ) at the viewing distance of 45 cm with a luminance of 196 cd/m2 . These were presented briefly ( 1 flash = 16 . 7msec ) , centred at 5 . 7dva to the left or right ( randomly across trials ) of a central fixation cross on an otherwise black screen . Auditory stimuli were presented via small speakers placed either side of the screen . These consisted of short beeps: 440Hz tones in a Gaussian temporal envelope of σ = 21msec . To reduce the reliability of the auditory stimuli , these beeps were embedded in continuous white noise [4] . As in previous studies , sequences of flashes and beeps were temporally aligned [14] , see Fig 9a . The spacing between events ( the ISI ) was varied as a function of age group , as determined by pilot work . This was done to roughly equate task difficulty across groups such that floor or ceiling effects were avoided: pilot work showed that a fixed ISI across groups resulted in floor effects for the youngest group ( such that the number of perceived events did not systematically increase as a function of the true number of events ) and / or ceiling effects in the adult group ( no response errors ) . For children in school years 1–3 ( infant school; age 4–7 years ) beeps and / or flashes were spaced by an ISI of 200msec . For junior school children ( school years 4–7; age 7–11 years ) the ISI was 167msec and for adults it was 117msec . All participants were given detailed instructions , and completed 8 practice trials in which they reported flashes ( 4 trials ) or beeps ( 4 trials ) . When counting flashes , subjects were told to ignore any beeps and vice versa . To help with motivation and concentration , participants were told that they needed to help Stinker the dog count beeps or flashes in order to get his treats . At the start of each block of experimental trials , Stinker appeared on the screen and instructed the participant to ‘count the flashes’ or ‘count the beeps’ . Each trial began with an ‘F’ or a ‘B’ presented at the screen’s centre to remind participants of the current task . To ensure fixation , participants were required to use the mouse to click this letter . The letter then changed to a green fixation cross and the sequence of flashes and / or beeps was presented . Each sequence consisted of 0–3 beeps and 0–3 flashes , such that the trial could be uni-modal ( only beeps or only flashes ) , bi-modal congruent ( equal number– 1 , 2 , or 3 –of flashes and beeps ) or bi-modal conflict ( the number of beeps and flashes differed by 1 or 2 ) . Uni-modal and bi-modal trials were randomly intermingled , but trials were blocked by focal modality ( i . e . report flashes , or report beeps ) . Participants gave their response on each trial by selecting the appropriate number on the keyboard ( 1–9 ) ; they were not told the maximum or minimum number of possible beeps or flashes . To keep the task duration within the concentration span of the child participants ( approximately 20 minutes , based on pilot work ) , infant school children completed 140 trials ( 2 modalities: judging beeps or flashes ) x 10 conditions ( 3 uni-modal , 7 bi-modal ) x 7 repetitions . Junior school children completed 8 repetitions ( 160 trials ) and adults completed 12 repetitions ( 240 trials ) . We report data from 76 observers ( 60 children , 16 adults ) . A further 5 children from the 4–6 age group were excluded who failed to complete the task and / or could not reliably count up to 3 . To check for counting ability / task comprehension , we used leave-one-out cross validation to compare regression models for each observer’s data , to ensure that the reported number of events across uni-modal and bi-modal congruent trials increased significantly as a function of the true number of events . Children were a priori divided into four age groups , by splitting the infant and junior school children at the midpoint of each age range , such that all children within a group were given the same stimulus set ( i . e . the same ISI ) . The resultant 5 groups were ( i ) ‘4–6 Years’: Range: 4 years 9 months to 6 years 3 months , n = 9 , 6 males ( ii ) ‘6–7 Years’: Range 6yrs 5m to 7yrs 8m , n = 11 , 7 males ( iii ) ‘8–9 Years’: 7yrs 9m to 9 yrs 8m , n = 19 , 9 males ( iv ) ‘10–11 Years’: 9yrs 11m to 11yrs 5m , n = 22 , 8 males and ( v ) ‘Adults’: Range 18–41 years , n = 16 , 8 males ) . The study was approved by the ethics committee at the University of Southampton and all participants gave informed consent . Parents / guardians gave consent on behalf of their children and children also provided consent on the day of the experiment .
|
To complete everyday activities , such as judging where or when something occurred , we combine information from multiple senses such as vision and audition . In adults , this merging of information is optimal: more reliable sensory estimates have more influence ( higher weight ) in the combined , multisensory estimate . Multisensory integration can result in illusions: if a single visual flash ( e . g . a bright disk appearing briefly on a screen ) occurs at the same time as two beeps , we sometimes perceive two flashes . This is because auditory information is generally more reliable than vision for judging when things happen; it dominates our audio-visual percept for temporal tasks . Previous work suggests that children don’t combine information from different senses in this adult-like way until around 10 years . To investigate this further , we asked children and adults to report the number of visual flashes or auditory beeps when these were presented simultaneously . Surprisingly , all children used appropriate sensory weights: audition—the more reliable signal—tended to dominate perception , with less weight given to vision . However , children didn’t show the adult-like reduction in uncertainty until around 8–10 years . Before that age , they switched between using only auditory or only visual information on each trial .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"children",
"acoustics",
"education",
"visual",
"signals",
"sociology",
"social",
"sciences",
"neuroscience",
"age",
"groups",
"probability",
"distribution",
"mathematics",
"vision",
"families",
"schools",
"probability",
"theory",
"people",
"and",
"places",
"physics",
"psychology",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"sensory",
"perception",
"physical",
"sciences",
"acoustic",
"signals"
] |
2016
|
The Development of Audio-Visual Integration for Temporal Judgements
|
Animals harbor specialized neuronal systems that are used for sensing and coordinating responses to changes in oxygen ( O2 ) and carbon dioxide ( CO2 ) . In Caenorhabditis elegans , the O2/CO2 sensory system comprises functionally and morphologically distinct sensory neurons that mediate rapid behavioral responses to exquisite changes in O2 or CO2 levels via different sensory receptors . How the diversification of the O2- and CO2-sensing neurons is established is poorly understood . We show here that the molecular identity of both the BAG ( O2/CO2-sensing ) and the URX ( O2-sensing ) neurons is controlled by the phylogenetically conserved SoxD transcription factor homolog EGL-13 . egl-13 mutant animals fail to fully express the distinct terminal gene batteries of the BAG and URX neurons and , as such , are unable to mount behavioral responses to changes in O2 and CO2 . We found that the expression of egl-13 is regulated in the BAG and URX neurons by two conserved transcription factors—ETS-5 ( Ets factor ) in the BAG neurons and AHR-1 ( bHLH factor ) in the URX neurons . In addition , we found that EGL-13 acts in partially parallel pathways with both ETS-5 and AHR-1 to direct BAG and URX neuronal fate respectively . Finally , we found that EGL-13 is sufficient to induce O2- and CO2-sensing cell fates in some cellular contexts . Thus , the same core regulatory factor , egl-13 , is required and sufficient to specify the distinct fates of O2- and CO2-sensing neurons in C . elegans . These findings extend our understanding of mechanisms of neuronal diversification and the regulation of molecular factors that may be conserved in higher organisms .
The capacity of the nervous system to sense and respond to fluctuations in the external and internal environment is essential for homeostasis and survival . Neuronally controlled homeostatic buffering is delivered through cellular and systemic physiological adjustments and by seeking optimal environmental conditions through behavioral strategies [1]–[4] . A crucial homeostatic capacity of animals is the ability to sense and respond to changes in concentration of the respiratory gases oxygen ( O2 ) and carbon dioxide ( CO2 ) [5] , [6] . O2 is essential for the generation of energy in the form of adenosine triphosphate ( ATP ) ; however , O2 also exerts toxicity through the production of reactive oxygen species ( ROS ) [1]–[4] , [7] . CO2 is a by-product of oxidative metabolism and prolonged exposure leads to acidosis [5] , [6] , [8] . CO2 is also an environmental cue used in host- and mate- finding and can initiate both aversive or attractive behaviors [9]–[11] . The evolution of mechanisms required to sense and respond to O2 and CO2 is therefore paramount for survival . In Drosophila , specific sensory systems respond to external O2 levels [1] , [7] , [12] . In addition , Drosophila uses specialized olfactory and gustatory neurons to detect CO2 changes via specialized chemosensory receptors called Gr21a/Gr63a [9] , [13] . In humans , O2 , CO2 and pH levels are monitored by specific regions of the brainstem and by specialized neurosecretory glomus cells of the carotid body [14] , whereas in non-human mammals CO2 is also sensed by specific olfactory neurons that target the necklace glomeruli in the olfactory bulb via the guanylyl cyclase GC-D [15] . It is poorly understood how the specification of such specialized sensory neurons is regulated . However , recent work in Drosophila has shown that epigenetic mechanisms play an important role [16] . Respiratory gas sensing is a crucial modality for Caenorhabditis elegans whose natural environment , such as rotting fruit and compost , can have wide ranges of O2 and CO2 levels [17] . Previous work has shown that in the laboratory , worms have a behavioral preference for 5%–10% O2 and are exquisitely sensitive to minor changes in O2 concentration [18] , [19] . In addition , worms mount avoidance responses to CO2 levels above 0 . 5% [4] , [11] . Of the 302 neurons in the C . elegans nervous system , at least six neurons are specifically dedicated to the detection and response to changes in O2 and CO2 levels . These include the BAGL/R , URXL/R , AQR and PQR neurons . The BAG neurons are the primary CO2 sensors and they also respond to decrease in O2 concentration [11] , [20]–[22] . The URX , AQR and PQR neurons are specialized for responding to increasing O2 concentrations [20] . In C . elegans , members of the guanylyl cyclase family of proteins are crucial factors required for O2 and CO2 sensing . Pioneering work revealed that the soluble guanylyl cyclases ( sGCs ) GCY-35 and GCY-36 mediate high O2 avoidance behavior via the URX , AQR and PQR neurons and that GCY-35 directly binds to molecular O2 [18] . In contrast , the sGCs , GCY-31 and GCY-33 function in the BAG neurons to sense decreases in O2 [20] . Recent work found that the membrane-bound receptor-type guanylyl cyclase GCY-9 acts specifically in the BAG neurons to mediate CO2 avoidance behavior [21] . Other molecules such as the Phe-Met-Arg-Phe-NH2 ( FMRF-amide ) -related peptides ( FLP-8 , FLP-13 , FLP-17 and FLP-19 ) are expressed in either a subset or all of the O2- and CO2-sensing neurons; however their precise molecular functions in O2 and CO2 sensing are not known [23] , [24] . Neuronal specialization within the O2/CO2-sensing system in C . elegans is an excellent model to study the control of neuron diversity . The O2-sensing ( URX , AQR and PQR ) and O2/CO2-sensing ( BAG ) neurons have overlapping and non-overlapping patterns of guanylyl cyclase and neuropeptide expression , which are reflected in their related , albeit distinct functionalities [20] . At present , it is unclear how the expression of these molecules is restricted to certain parts of the O2/CO2-sensing nervous system , and how such restrictions coordinate neuronal fate and function . Here , we have identified the Sox transcription factor EGL-13 as an important regulator of the O2 and CO2-sensing neuron cell fate decision . EGL-13 is required for the expression of distinct proteins required for sensing both O2 and CO2 and as such , egl-13 mutant animals are unable to mount behavioral responses to changes in O2 and CO2 . We found that the expression of EGL-13 is controlled by ETS-5 in the BAG neurons and by AHR-1 in the URX neurons , and acts in partially parallel pathways with these factors to drive neuronal fate . Finally , we found that EGL-13 is sufficient to drive O2- and CO2-sensing cell fates in certain cellular contexts . Therefore , EGL-13 is a core regulatory factor that is both required and sufficient to drive O2- and CO2-sensing neuron specification in C . elegans . As EGL-13 is a member of the SoxD family of transcription factors , we anticipate that the regulatory relationships described here will provide a paradigm for the control of neuronal fate specification by Sox proteins in other cellular contexts .
In order to identify molecules and pathways important for O2- and CO2-sensing neuron specification , we have taken a forward genetics approach in C . elegans . We isolated four independent allelic mutations ( rp14 , rp22 , rp23 and rp26 ) that affect the expression of terminal differentiation markers in the O2 and/or CO2-sensing neurons ( Figure 1 and Table S1 ) Mutant hermaphrodites of each of these alleles are severely egg-laying defective ( Egl ) and form a bag-of-worms where embryos hatch inside the mother ( Figure S1 ) . We investigated their vulval phenotype and found that the anchor cell fails to fuse with the uterine seam cell , causing a blockage of the uterus and the resultant Egl phenotype ( Figure S1 ) . This anchor cell fusion defect is reminiscent of that observed in egl-13 ( ku194 ) mutant animals [25] which we found to also exhibit defects in O2 and CO2 reporter expression ( Figure 1 and Figure S1 ) . Subsequent Sanger sequencing of rp14 , rp22 , rp23 and rp26 revealed genetic lesions in the egl-13 locus ( Figure 1A ) . egl-13 encodes the C . elegans ortholog of the HMG-domain-containing SoxD family of transcription factors that has no previously reported role in the worm nervous system . The BAG , URX , AQR and PQR neurons in C . elegans are required for sensing and responding to fluctuations of O2 and CO2 levels in the environment [11] , [18] , [19] . Distinct batteries of genes are expressed in these neurons that are predicted to provide the optimal functionality required for O2 and CO2 sensing , however the role of only a subset of these genes has been analyzed in detail [20] , [21] , [26] . We used fluorescent reporter constructs to monitor expression of these gene batteries to understand how egl-13 controls O2 and CO2-sensing neuron cell fate ( Figure 1 ) . We analyzed the expression of guanylyl cyclases ( gcy-9 , gcy-31 , gcy-33 , gcy-35 and gcy-36 ) and Phe-Met-Arg-Phe-NH2 ( FMRF-amide ) -related peptides ( flp-8 , flp-13 , flp-17 and flp-19 ) that are all terminal differentiation genes expressed in all or a subset of O2 and CO2-sensing neurons [18] , [19] , [21] , [24] . We crossed these reporter transgenes into egl-13 mutant animals ( ku194 allele ) and found that none of the reporters were properly expressed in egl-13 mutants ( Figure 1 and Table S1 ) . We also found similar effects in the four egl-13 mutant alleles we isolated ( rp14 , rp22 , rp23 and rp26 ) ( Figure S1C ) . We noticed that some of the reporters were exquisitely sensitive to egl-13 loss whereas others exhibited partially penetrant defects ( Table S1 ) . This suggests that the expression of some terminal differentiation factors are under the collaborative control of additional factors that are able to compensate for the loss of egl-13 . The BAG and URX neurons are derived from the AB lineage and are posterior sisters of other neurons that have distinct fates [27] ( Figure S2 ) . We therefore asked whether egl-13 is also required for the specification of the sister cells of BAG or URX . We crossed egl-13 ( ku194 ) mutant animals into fluorescent reporter strains for the SMDV , zfIs2 ( lgc-55::mCherry ) and CEPD , vtIs1 ( dat-1::gfp ) , sister cells for BAG and URX neurons respectively . We found that the expression of these reporters were unaffected by loss of egl-13 suggesting a specific role for egl-13 in the posterior branch of these lineages ( Table S1 and Figure S2 ) . Taken together , we conclude that egl-13 controls the expression of the distinct O2- and CO2-sensing neuron terminal gene batteries that distinguish them from lineage-related neurons . To monitor egl-13 expression , we generated two promoter-driven fluorescent reporters ( egl-13prom1::mCherry and egl-13prom1::gfp ) that contain 3 . 5 kb of egl-13 upstream sequence ( Figure 2A and Figure S3 ) . Expression is first detected in 4 neuronal cells at around 350 min post-fertilization , which is the time at which the BAG and URX neurons are born ( Figure S3 ) . Expression is restricted to these 4 neurons during embryogenesis ( Figure S3 ) . At the first larval stage , egl-13 expression is observed in the BAG and URX neurons plus occasionally in a small number of unidentified cells in the head and tail ( including the AQR and PQR neurons ) ( Figure S3 ) . Later during larval development , egl-13 expression is observed in body wall muscle and vulval cells ( data not shown ) . Neuronal expression is restricted to the O2 and CO2-sensing neurons in the adult ( Figure 2A ) . Using the 3 . 5 kb egl-13 promoter ( egl-13prom1 ) we transgenically expressed egl-13isoformAcDNA in egl-13 ( ku194 ) mutant animals and were able to rescue both the defect in O2 and CO2-sensing neuron fate marker expression and the Egl phenotype ( Figure 2B–2C , Figure S4 and data not shown ) . To confirm that egl-13 acts cell autonomously to control O2 and CO2-sensing neuron fate , we used neuron-specific promoters to drive egl-13isoformAcDNA expression in the BAG or URX neurons ( Figure 2D ) . We found that indeed neuron-specific expression of egl-13 rescued the O2 and CO2-sensing neuron fate defect of egl-13 ( ku194 ) mutant animals ( Figure 2D ) . Therefore , we conclude that egl-13 acts autonomously in the BAG and URX neurons to direct their fate . The egl-13 gene has 4 predicted isoforms , all of which contain the same HMG DNA/protein binding domain , however they each have varying lengths of amino terminal tail . Such tails in SoxD proteins can cooperate with other factors to control gene expression [28] . We therefore tested whether the long N-terminal region of EGL-13isoformA is required for its rescuing ability . We used egl-13prom1 to drive EGL-13isoformD ( lacking 157 amino acids of the N-terminal tail of isoformA ) in egl-13 ( ku194 ) animals and found that it fully rescued the defect in O2 and CO2-sensing neuron fate marker expression and the Egl phenotype ( Figure S4 and data not shown ) . Thus , the EGL-13 N-terminal region is not required for its roles in vulval cell nor O2 and CO2-sensing neuron specification . We next asked whether SoxD proteins play specific roles in these decisions by attempting to rescue the egl-13 mutant defects with the SoxB family member , sox-2 . We expressed sox-2 cDNA under the control of egl-13prom1 in egl-13 ( ku194 ) mutant animals . We found that sox-2 is unable to rescue O2 and CO2-sensing neuron fate marker expression ( Figure 2E ) . These data indicate that the SoxD HMG domain plays a specific role in the specification of O2 and CO2-sensing neuron fate in C . elegans . We have shown that egl-13 is expressed throughout the life of the worm in the O2 and CO2-sensing neurons; and is required to induce terminal differentiation features . To ask whether egl-13 is required continuously to maintain the expression of the terminal gene battery of these neurons , we sought to postdevelopmentally remove egl-13 gene activity . egl-13 gene activity could not be removed by RNA-mediated interference in an RNAi sensitized background ( data not shown ) and there are no temperature-sensitive alleles of egl-13 available . Instead , we generated animals that lack endogenous EGL-13 protein but express heat-shock inducible egl-13 cDNA from an extrachromosomal array under the control of the hsp-16 . 2 promoter ( Figure 3 ) . We focused our analysis on the URX neurons and found that the loss of gcy-33prom::gcy-33::gfp reporter expression in egl-13 ( ku194 ) worms could be rescued through heat-shock induction of egl-13 during mid-larval stages ( Figure 3A ) . This indicates that O2-sensing neurons generated during embryogenesis persist in an egl-13-responsive state . These neurons are , therefore , not converted into another fate when egl-13 is lost; however , they do not acquire the terminal O2-sensing neuron differentiation program . When egl-13 activity was supplied transiently , through removal of heat-shock stimulus , we observed a gradual loss of reporter expression during adulthood in the URX neurons ( Figure 3A ) . Therefore , egl-13 gene activity is continuously required to maintain URX cell fate . To ask whether misexpression of egl-13 in other neurons is sufficient to induce O2 and CO2 terminal fate we expressed egl-13 under the control of an early neuronal promoter ( Figure 3B ) . We found that egl-13 is indeed sufficient to induce expression of O2 and CO2 terminal fate markers in some cellular contexts ( Figure 3B ) . This suggests that egl-13 is not only required but also sufficient to induce O2 and CO2-sensing neuron fate in specific contexts , which is similar to previous studies of terminal selector genes [29]–[31] . The restricted induction we observed may be dependent on the embryonic time-point of induction or the expression of other unknown co-factors that are required for induction of O2 and CO2-sensing neuron fate . The crucial role for egl-13 in O2 and CO2-sensing neuron fate determination suggested that egl-13 mutant animals would be defective in O2 and CO2 sensing . We applied three behavioral paradigms that have been previously reported to be specific to either one of these neuron classes: BAG neurons modulate the animals' locomotion speed in response to an oxygen downshift from 21% O2 towards 10% O2 ( Figure 4A , 4E ) [20] . In addition , BAG neurons detect increases in CO2 concentrations , which trigger reorientation movements ( omega turns ) ( Figure 4G ) [11] , [21] . URX neurons modulate the animals' locomotion speed in response to O2 upshifts towards 21% O2 ( Figure 4A , 4F ) [20] . We applied these behavioral assays to test how BAG and URX neurons are functionally affected in egl-13 mutants . We tracked animals in a chamber without food , in an air-flow that switched between 21% O2 and 10% O2 , or between 0% CO2 and 1% CO2 . In contrast to wild-type animals , egl-13 ( ku194 ) mutant animals do not slow their locomotion in response to O2 upshift or downshift ( Figure 4A , 4C , 4E , 4F ) . We found that egl-13 ( ku194 ) mutants are also defective in CO2 sensing since they fail to slow or perform omega turns in response to CO2 ( Figure 4G ) . O2 and CO2 behavioral defects of egl-13 ( ku194 ) mutants are fully rescued when egl-13 cDNA is resupplied under the control of egl-13prom1 ( Figure 4D–4G ) . These data confirm that egl-13 is crucial for the specification and function of O2 and CO2-sensing system in C . elegans . One of the egl-13 mutant alleles retrieved from our screen was a promoter deletion mutant ( rp23 ) . The rp23 deletion removes 1128 bp of egl-13 promoter from −1700 to −572 upstream of the translational start site ( Figure 1A and Figure 5A ) . Intriguingly , the rp23 mutation affects terminal marker expression in the URX but not BAG sensory neurons and is mostly defective in URX and less affected in BAG regulated behaviors ( Figure 4B , 4E–4G; Figure S4C; and Table S1 ) . This suggests that the rp23 promoter deletion removes element ( s ) required to drive egl-13 in the URX neurons while leaving the BAG-specific element ( s ) intact . To identify which upstream factors drive expression of egl-13 in the molecularly and functionally distinct BAG and URX neurons we performed promoter deletion analysis , using the 3 . 5 kb upstream element ( egl-13prom1 ) as a template . We generated transgenic worms expressing truncated versions of egl-13prom1 driving mCherry or gfp protein and focused our expression analysis on BAG and URX regulation ( Figure 5A ) . A 900 bp fragment ( egl-13prom3 ) , which includes 360 bp corresponding to the 3′ end of the rp23 deletion , drove expression in BAG and URX . However , a 691 bp fragment ( egl-13prom4 ) , which lacks the missing region in the rp23 deletion , only drove expression in the BAG neurons . Therefore , an important element required for egl-13 expression specifically in the URX neurons lies within the 200 bp region included in egl-13prom3 . Bioinformatic analysis of this region revealed that there are two conserved motifs that are potential binding sites for EGL-13/SOX5 itself and AHR-1 , an aryl hydrocarbon receptor bHLH protein . Interestingly , ahr-1 was previously shown to be required for the expression of some URX terminal fate markers [32] . Site-directed mutagenesis of the predicted AHR-1 binding site significantly reduced egl-13prom1::mCherry expression and a subsequent mutation in the putative EGL-13 binding site further reduced expression ( Figure 5A ) . This suggests that both AHR-1 and EGL-13 regulate egl-13 expression . To test this , we crossed egl-13prom1mCherry/gfp-expressing animals into egl-13 ( ku194 ) and ahr-1 ( ia3 ) mutants and found that URX expression was reduced in both cases ( Figure 5B ) . Therefore , AHR-1 and EGL-13 both contribute to the control of egl-13 expression in the URX neurons . To identify the regulatory module ( s ) that control egl-13 expression in the BAG neurons , we continued to dissect the egl-13 promoter . We identified a 432 bp region ( egl-13prom5 . 1 ) , immediately upstream of the ATG codon , which is sufficient to drive expression in the BAG neurons . Intriguingly , we found two conserved ETS-5/Pet1 binding sites in this region ( Figure 5 ) . Previous work identified ETS-5 as a crucial factor required for the specification of the BAG neurons , suggesting that ETS-5 may regulate egl-13 expression in these neurons [26] , [33] . We used site-directed mutagenesis to eliminate the ETS-5/Pet1 binding sites individually and in combination , and found that when both ETS-5/Pet1 binding sites are mutated the expression of egl-13 is abrogated in the BAG neurons ( Figure 5A ) . This suggests that ETS-5 directly regulates the expression of egl-13 in the BAG neurons via conserved binding sites . We crossed the ets-5 ( tm1734 ) mutant into the egl-13prom1::mCherry strain and indeed found that BAG expression was affected ( Figure 5B ) . In addition , we found that egl-13 can regulate its own expression in the BAG neurons independently of ets-5 via an , as yet , unidentified mechanism ( Figure 5B ) . Taken together , these data indicate that control of egl-13 expression is coordinated by two independent regulatory mechanisms . In the O2-sensing URX neurons , egl-13 expression is predominantly regulated by AHR-1 ( Figure 5C ) . In contrast , an independent promoter module controlled by ETS-5 regulates egl-13 expression in the O2/CO2-sensing BAG neurons ( Figure 5C ) . In addition , egl-13 is able to autoregulate in both the URX and BAG neurons ( Figure 5C ) . Previous work has identified ets-5 and ahr-1 as regulators of BAG and URX specification respectively [26] , [32] , [33] and we have shown that these factors are predominantly required to drive egl-13 expression in these cells . To understand how these factors function together to coordinate BAG and URX specification , we analyzed the expression of terminal fate markers in single and double mutant combinations , where appropriate . We analyzed three URX markers ( flp-8::gfp , flp-19::gfp and gcy-33prom::gcy-33::gfp ) and six BAG markers ( flp-13::gfp , flp-17::gfp , flp-19::gfp , gcy-9::gfp , gcy-31::gfp and gcy-33prom::gcy-33::gfp ) and compared the effect of the individual loss of egl-13 , ets-5 and ahr-1 ( Figure S5 ) . The first observation from this analysis was that the expression of a subset of terminal differentiation markers is completely dependent on egl-13 and one of the other factors acting in a linear pathway . For example , we find that BAG expression of flp-13::gfp and flp-19::gfp is almost 100% affected in both the egl-13 and ets-5 single mutants ( Figure S5 ) . This suggests that for these markers egl-13 and ets-5 act in the same pathway to drive marker expression . In contrast , expression of gcy-9::mCherry is completely dependent on ets-5 with egl-13 playing a minor role in its regulation ( Figure S5 ) . At the other end of the spectrum , ets-5 and egl-13 are minimally required to drive gcy-31::mCherry expression in the BAG neurons suggesting other factor ( s ) control the expression of this terminal fate marker ( Figure S5 ) . Taken together , these data indicate that egl-13 and ets-5 act in partially parallel pathways to drive BAG cell fate and that other unknown factors possibly act in a combinatorial manner to drive specific aspects of BAG fate . We also observed differential effects of egl-13 loss with URX terminal fate markers . Expression of the flp-8::gfp reporter is partially affected by single loss of egl-13 and ahr-1 , whereas loss of both genes totally abrogates expression , suggesting that egl-13 and ahr-1 act in parallel pathways to regulate flp-8::gfp expression ( Figure S5 ) . However , in the case of flp-19::gfp , loss of egl-13 causes complete loss of expression and ahr-1 plays a minor role in its regulation ( Figure S5 ) . To further investigate the regulatory relationship between egl-13 , ets-5 and ahr-1 we analyzed how they affect the expression of each other . We have already shown that ets-5 positively regulates the expression of egl-13 in the BAG neurons ( Figure 5B ) . In a reciprocal experiment , we found that ets-5::gfp expression is unaffected in egl-13 ( ku194 ) mutant animals ( Figure 5B ) . These data and other work [26] , [33] suggest that ets-5 acts upstream and in parallel to egl-13 to direct BAG cell fate ( Figure 5C ) . In addition , we found that egl-13 is able to regulate its own expression in the BAG neurons , in parallel to ets-5; however , the mechanistic basis of this regulation is unclear ( Figure 5B ) . In the URX neurons , we found that egl-13 and ahr-1 regulate the expression of each other in addition to having autoregulatory capabilities ( Figure 5B , C and Figure S5 ) . Our studies have elucidated a novel function for egl-13 , the SoxD homolog , in the specification of distinct classes of O2 and CO2 sensory neurons in C . elegans . We show that egl-13 is expressed in the O2- and CO2-sensing neurons and acts cell-autonomously to regulate their distinct cell fates . We further show that egl-13 is continuously expressed in the O2- and CO2-sensing system to maintain the expression of terminal features of these neurons . In certain cellular contexts , egl-13 is also sufficient to induce O2- and CO2-sensing neuron cell fate . We found that the regulatory inputs controlling the expression of egl-13 in the O2- and CO2-sensing system are mechanistically distinct . Independent regulatory modules control egl-13 expression in the BAG neurons ( CO2 and O2 downshift sensors ) versus the URX neurons ( O2 upshift sensors ) . Interestingly , we found that egl-13 expression in the BAG neurons is controlled by the ETS-5 transcription factor via conserved ETS binding sites . In contrast , in the URX neurons , egl-13 expression is controlled by the bHLH transcription factor AHR-1 via a conserved AHR1 binding site . The influence EGL-13 exerts on the expression of the terminal gene batteries of the distinct O2- and CO2-sensing neurons is diverse . Particular factors are exquisitely sensitive to loss of egl-13 , whereas others are only partially affected . These findings suggest that alternative unknown modes of regulation are in place to ensure that particular molecules are faithfully expressed in the O2 and CO2 sensory neurons , which work in conjunction with and/or in parallel to egl-13 . Sox transcription factors have diverse functions during development and play crucial roles in regulating neuronal fate [34]–[37] . In addition , Sox proteins act at different levels to preselect neuronal genes in embryonic stem cells and to direct the activation of these genes in neuronal precursors and fully differentiated neurons [38] . Here we describe a novel role for EGL-13 , the SoxD transcription factor in C . elegans , in driving the specification of different but related sensory neuron identities . Closely related orthologs of EGL-13 are found in vertebrates , some of which are expressed in sensory neurons [39] , therefore; SoxD proteins may have a previously unrecognized conserved function in the specification of gas-sensing neurons in higher organisms .
Strains were grown using standard growth conditions on NGM agar at 20°C on Escherichia coli OP50 [8] , [40] . Transgenic animals were created according to [41] . Strain information is detailed in Table S2 . In all screens , animals were mutagenized with EMS ( ethyl methanesulfonate ) according to standard protocols [42] . Worms were incubated at 25°C at all times . In the manual screens , 5 parental ( P0 ) mutagenized animals were placed in each of 10 founder plates . Three days later , 400 F1 progeny of the mutagenized P0 animals were singled . Their ensuing F2 progeny were screened under a fluorescence stereomicroscope . In the automated worm sorter screen , around 100 , 000 synchronized larval stage L4 animals were mutagenized with EMS , the following day the P0 young adult animals were bleached and their F1 progeny synchronized at larval stage L1 by starvation ( approximately 1 , 000 , 000 animals ) . F1 animals were grown to the young adult stage , bleached and their F2 progeny synchronized at larval stage L1 by starvation ( approximately 10 , 000 , 000 animals ) . The F2 progeny were grown until larval stage L4 and 10% of the population ( approximately 1 , 000 , 000 ) was passed through a COPAS biosorter ( Karolinska Institute , Stockholm , Sweden ) . Reporter gene constructs were generated by PCR amplifying promoter elements and cloning into the pPD95 . 75-mCherry and gfp vectors ( Fire Vector Kit ) . Mutagenesis was performed using the QuikChange II XL Site-Directed Mutagenesis Kit ( Stratagene ) . Rescue constructs were generated by cloning promoter and cDNA sequences into the pPD49 . 26 expression vector ( Fire Vector Kit ) . Constructs were injected into young adult hermaphrodites as either simple arrays ( gcy-33prom::gcy-33::gfp ( 50 ng ul−1 ) and pRF4 ( 50 ng ul−1 ) as injection markers ) or as complex arrays using 1–10 ng ul−1 of linearized plasmid , 150 ng ul−1 of PvuII-digested bacterial genomic DNA and myo-2prom::dsRed ( 3–5 ng ul−1 ) , elt-2prom::gfp ( 3–15 ng ul−1 ) as injection markers . Animals were transferred without food to 14 cm NGM assay plates containing a cut out arena of Whatman filter paper soaked in 20 mM CuCl2 to prevent them from leaving a 56 mm×56 mm center area . Sixty to seventy animals were used in a single experiment and starved for one hour prior to examination . Each experiment was carried out three times , except for wild-type , which was performed six times . A custom-made transparent plexiglass device with a flow arena of 60 mm×60 mm×0 . 7 mm was placed onto the assay arena and animals were accustomed to a gas flow of 100 ml/min containing 21% ( v/v ) oxygen for 5 minutes . During the assays animals were exposed for 6 minutes to 21% O2 before and after a 6 minute stimulus interval of either 10% O2 or 1% CO2 ( +21% O2 ) . All gas mixtures were balanced with N2 . Gases were mixed with a static mixing element connected to mass flow controllers ( Vögtlin Instruments ) that were operated by LabView software . Recordings of freely behaving animals illuminated with flat red LED lights were made at 3 fps on a 4 megapixel CCD camera ( Jai ) using Streampix software ( Norpix ) . Movies were analyzed by MatLab-based image processing and tracking scripts as previously described [43] , [44] . The resulting trajectories were used to calculate instantaneous speed during continuous forward movements ( 1 second binning ) . Omega turns were detected based on characteristic changes in object eccentricity and their frequency was calculated in 15 second bins . For quantifications , relative speed changes were calculated between representative intervals of 120 seconds before ( basal level ) and 4 seconds after the stimulus , capturing the minimum speed levels ( 4–8 seconds post stimulus ) . Data were normalized to the basal level . Changes in omega turn frequency were calculated between representative intervals of 180 seconds before ( basal level ) and 60 seconds after the stimulus , to capture the maximum rise phase ( 55–115 seconds post stimulus ) . Two transgenic lines for hsp-16 . 2::egl-13 were used for the heat-shock experiments . For the rescue and maintenance experiments , third larval stage ( L3 ) worms were heat shocked at 37°C two times for 30 min . After heat shock , worms were kept at 25°C overnight and then transferred to 15°C for 2 days . Worms were mounted on 5% agarose on glass slides and images were taken using an automated fluorescence microscope ( Zeiss , AXIO Imager M2 ) and MicroManager software ( version 3 . 1 ) . Neurons were given a numerical value according to their expression levels . Wild-type expression scored 1 , decreased expression scored 0 . 5 and abolished expression scored 0 . Percentage of GFP expressing animals was then correlated to the theoretical maximum score using the equation below . The Jaspar program ( http://jaspar . genereg . net/ ) was used to predict the transcription factor binding sites in the egl-13 upstream regulatory sequence . Statistical analysis was performed in GraphPad Prism 5 using one-way ANOVA with Newman-Keuls Multiple Comparison Test . Values are expressed as mean ± s . d . Differences with a P value<0 . 05 were considered significant . For the behavioral assays statistical significance was determined using one-way ANOVA with Bonferroni's Multiple Comparison Test .
|
During the development of an organism , certain neurons are programmed to perform specific tasks . For example , motor neurons coordinate locomotion and sensory neurons recognize specific environmental cues . The molecular mechanisms that generate specific neuronal classes are not fully understood . We investigated mechanisms that control the development of two distinct classes of neurons that are required for the nematode Caenorhabditis elegans to sense the respiratory gases O2 or CO2 . In this study , we identified and characterized a conserved transcription factor , egl-13 , that is required for the development of both of these classes of neurons . egl-13 is related to the SoxD family of transcription factor proteins in vertebrates . We found that egl-13 controls the production of specific proteins that provide these cells with the ability to sense both O2 and CO2 . Further , we found that egl-13 works in conjunction with two additional factors , ahr-1 and ets-5 , to regulate this developmental decision . This work provides new insight into how transcriptional regulatory networks specify different but related neuronal identities and provides a platform for future studies to understand how neuronal diversity is generated .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"neuroscience",
"developmental",
"neuroscience",
"behavioral",
"neuroscience",
"biology",
"neuroscience"
] |
2013
|
EGL-13/SoxD Specifies Distinct O2 and CO2 Sensory Neuron Fates in Caenorhabditis elegans
|
Myopia ( nearsightedness ) is the most common eye disorder , which is rapidly becoming one of the leading causes of vision loss in several parts of the world because of a recent sharp increase in prevalence . Nearwork , which produces hyperopic optical defocus on the retina , has been implicated as one of the environmental risk factors causing myopia in humans . Experimental studies have shown that hyperopic defocus imposed by negative power lenses placed in front of the eye accelerates eye growth and causes myopia , whereas myopic defocus imposed by positive lenses slows eye growth and produces a compensatory hyperopic shift in refractive state . The balance between these two optical signals is thought to regulate refractive eye development; however , the ability of the retina to recognize the sign of optical defocus and the composition of molecular signaling pathways guiding emmetropization are the subjects of intense investigation and debate . We found that the retina can readily distinguish between imposed myopic and hyperopic defocus , and identified key signaling pathways underlying retinal response to the defocus of different signs . Comparison of retinal transcriptomes in common marmosets exposed to either myopic or hyperopic defocus for 10 days or 5 weeks revealed that the primate retina responds to defocus of different signs by activation or suppression of largely distinct pathways . We also found that 29 genes differentially expressed in the marmoset retina in response to imposed defocus are localized within human myopia quantitative trait loci ( QTLs ) , suggesting functional overlap between genes differentially expressed in the marmoset retina upon exposure to optical defocus and genes causing myopia in humans . These findings identify retinal pathways involved in the development of myopia , as well as potential new strategies for its treatment .
Myopia ( nearsightedness ) is an eye disorder characterized by blurred distance vision caused by negative refractive errors when the eye grows too long for its optical power . It is widespread [1] , and the prevalence has been increasing around the world at an alarming rate , reaching 80%–90% in several parts of Asia [2–5] . Because of the increasing prevalence , myopia is rapidly becoming one of the leading causes of vision loss in several parts of the world , as the excessive eye growth associated with it often leads to serious vision-threatening complications , such as myopic maculopathy , chorioretinal atrophy , retinal tears , retinal detachment , myopic macular degeneration , and glaucoma [1 , 6–8] . It is estimated that 4 . 8 billion people ( half of the world’s population ) will be affected by myopia by 2050 [9] , predicting an impending public health crisis . The World Health Organization designated myopia as one of five priority health conditions [10 , 11] . Developmental studies in humans and a wide variety of animals show that eyes grow into focus postnatally through a process called emmetropization . During emmetropization , the axial length of the eye grows to match its optical power , producing focused images on the retina; a mismatch between the optical power of the eye and its axial length leads to the development of hyperopia if eyes are too short for their optical power or myopia if eyes grow too long for their optical power . Visual experience is critical for this process , suggesting that postnatal eye growth is guided by visual feedback related to retinal defocus . Experimental studies in many animal species confirm the role of optical defocus in emmetropization , with observations that the eye can compensate for either imposed hyperopic or imposed myopic defocus by either increasing or decreasing its growth rate [12–26] . Myopia can be induced in species as diverse as primates , tree shrews , guinea pigs , mice , chickens , and fish by placing a negative lens in front of the eye and imposing hyperopic defocus on the retina [13 , 16 , 20 , 21 , 23 , 25–28] . Imposing myopic defocus on the retina with positive lenses , on the other hand , inhibits eye growth and produces a compensatory hyperopic shift in refraction in fish [28] , chickens [21 , 23] , and several mammalian species , including primates [13 , 16 , 18 , 20 , 25 , 26] . The balanced response to these two optical signals of opposite signs seems to be regulating eye growth and refractive development . Moreover , animal experiments in which the optic nerve was sectioned [29–31] and in which different parts of the retina were simultaneously exposed to the defocus of opposite signs [32–34] suggested that the process of emmetropization is largely controlled locally by the retina . A number of studies have explored the cellular and molecular mechanisms of emmetropization , and while these studies have hinted at the involvement of various factors and pathways [35–44] , the field has made only incremental gains and has been at a conceptual standstill for some time . Very little is actually known about the biological signaling that underlies the eye’s response to different signs of optical defocus , and the ability of the retina to detect the sign of defocus has been the subject of much controversy because of the lack of direct evidence for the underlying mechanism [19 , 45 , 46] . Essentially , two alternative models for the visual control of emmetropization have been proposed . In the first model , the rate of eye growth is regulated by the amount of retinal blur , regardless of the sign of defocus . In this model , more blur promotes more growth , and imposed hyperopic or myopic defocus could be compensated for appropriately because hyperopic eyes generally experience more defocus than myopic eyes , which experience some focus at near . If the retina indeed could not distinguish between myopic and hyperopic defocus , both imposed myopic and hyperopic defocus would be expected to affect expression of the same genes in the same direction . The second model suggests that postnatal eye growth is controlled by the sign of retinal defocus and that response to defocus is bidirectional , i . e . , myopic defocus actively reduces growth and hyperopic defocus actively increases growth . This could be controlled by one controller that modulates growth rate or two separate controllers that work independently to increase or decrease growth . In the latter case , emmetropization is achieved through the balance between the two independent outputs . If this model is true , then imposed myopic and hyperopic defocus would be expected to modulate the expression of the same genes in opposite directions or affect expression of different sets of genes . In this study , we analyzed whole-genome gene expression using massive parallel RNA sequencing ( RNA-seq ) in the retina of a New World primate , common marmosets ( Callithrix jacchus ) , exposed to either myopic or hyperopic defocus . We demonstrate that primate retina ( 1 ) can distinguish between hyperopic and myopic optical defocus , and ( 2 ) responds to the different signs of defocus by activation or suppression of largely distinct pathways . These data support the hypothesis that emmetropization is regulated bidirectionally by two separate retinal controllers: one that involves active stimulation of eye growth by hyperopic defocus and another that actively suppresses eye growth in response to myopic defocus . We refer to this mechanism as Bidirectional Emmetropization by the Sign of Optical Defocus ( BESOD ) . Furthermore , we identified key signaling pathways underlying the retinal responses to the defocus of different signs , which offer possible targets that can be manipulated pharmacologically to suppress myopia . These results establish the critical role of the retina in the control of postnatal eye growth and emmetropization , and provide a framework for the development of drugs that can be used to treat myopia and help prevent the blinding complications associated with it .
The eyes of common marmosets compensate for optical defocus imposed by either negative or positive lenses by increasing or decreasing axial growth and developing myopic or hyperopic refractive errors , respectively [15 , 16 , 25 , 26] . To investigate the retina’s molecular responses to different signs of defocus , we analyzed changes in the whole-genome transcriptome in the retina of common marmosets ( C . jacchus ) exposed to myopic or hyperopic defocus using +5D or −5D single vision contact lenses , respectively . We applied the lenses to the right eye of 10–11-week-old ( mean 74 ± 5 days ) marmosets for 10 days to examine the effect of the initial exposure to retinal defocus , or for 5 weeks to explore the retinal responses when measurable changes in eye growth and refractive state typically just begin to be detected [25] . The left eye served as a control and was fitted with a plano contact lens ( Fig 1 ) . Table 1 provides descriptive optometric data of the subjects under each condition . The lens-imposed retinal defocus ( the effective refractive error measured through the lens at the end of the rearing period ) provides a measure of the average retinal defocus experienced by the subjects in each group at the time of the tissue collection . It shows that the +5D lenses imposed myopic defocus and the −5D lenses imposed hyperopic defocus after 10 days ( early response ) and 5 weeks ( sustained response ) of lens wear . As expected from earlier studies [25] , there was little if any compensation for the imposed defocus observed for these treatment durations , as seen from the average interocular differences in spherical equivalent refractive error or vitreous chamber depth between the experimental and control eyes . Following lens treatment , retinae were collected and used to perform whole-genome gene expression profiling using RNA-seq . In the eyes treated with −5D lenses for 10 days , a total of 119 genes , organized into 6 genetic networks , were differentially expressed , compared with controls ( Fig 2A and 2B; S1 and S5 Tables; S1 Fig ) . Specifically , 87 genes were up-regulated and 32 genes were down-regulated . After 5 weeks of exposure to −5D lenses , 309 genes were differentially expressed compared to controls , and these genes were organized into 17 genetic networks ( Fig 2E and 2F; S2 and S6 Tables; S2–S4 Figs ) : 106 genes were up-regulated and 203 genes were down-regulated . In the +5D-lens-treated animals , 79 genes , organized into 4 genetic networks , were differentially expressed after 10 days ( Fig 3A and 3B; S3 and S7 Tables; S5 Fig ) , and 740 genes , organized into 25 genetic networks , were differentially expressed after 5 weeks of treatment ( Fig 3E and 3F; S4 and S8 Tables; S6–S10 Figs ) , compared with the control eyes . In the animals treated with +5D lenses , a total of 53 genes were up-regulated and 26 genes were down-regulated after 10 days , whereas 507 genes were up-regulated and 233 genes were down-regulated after 5 weeks of treatment . In terms of the numbers of differentially expressed genes , negative-lens-imposed hyperopic defocus induced a stronger response at the level of gene expression compared to the positive-lens-imposed myopic defocus during the first 10 days of treatment , while positive lenses elicited a substantially stronger response compared to negative lenses after 5 weeks of treatment . Gene ontology ( GO ) analysis revealed that optical defocus affects numerous cellular functions , with noticeable differences between marmoset eyes exposed to negative and positive lenses , as well as between treatment durations of 10 days and 5 weeks ( Fig 2C and 2G; S9 and S10 Tables; Fig 3C and 3G; S11 and S12 Tables ) . This initial observation was further reinforced by the analysis of canonical pathways affected by the positive and negative lenses ( Fig 2D and 2H; S13 and S14 Tables; Fig 3D and 3H; S15 and S16 Tables ) . We found that the early response ( 10 days ) to hyperopic defocus imposed by −5D lenses primarily involved pathways regulating glycogen degradation , ephrin and reelin signaling , as well as biosynthesis of spermine and choline ( Fig 2D; S13 Table ) , while the sustained response ( 5 weeks ) involved activation of ß-adrenergic signaling and suppression of cAMP-mediated signaling , protein kinase A signaling , calcium signaling , androgen signaling , and dopamine-DARPP32 feedback signaling , among several other pathways ( Fig 2H; S14 Table ) . On the other hand , the early response to myopic defocus imposed by +5D lenses primarily involved phenylalanine degradation and RANK , SAPK/JNK , NGF , and gap junction signaling ( Fig 3D; S15 Table ) , while the sustained response involved activation of EIF2 , Notch , JAK/Stat , oncostatin M , somatostatin receptor 2 , interleukin , CNTF , CREB , α-adrenergic , integrin , and ceramide signaling as well as suppression of apoptosis and aldosterone signaling ( Fig 3H; S16 Table ) . Collectively , these results demonstrate that optical defocus causes large-scale changes in gene expression controlling metabolism and cell signaling in the retina , and that there are substantial differences in the retinal response to hyperopic and myopic defocus . Taking into account the differences in GO functions and signaling pathways involved in the defocus response in the four experimental groups , we additionally performed a systematic comparison of cellular functions and canonical pathways affected in the different experimental groups ( Figs 4 and 5; S9–S20 Tables ) . This analysis revealed that each experimental condition leads to differential expression of a unique set of genes , with very little overlap between the datasets ( Fig 4A and 4C; Fig 5A and 5C ) . There was only an eight-gene overlap between the +5D/10-days group and the +5D/5-weeks group , with three genes ( PIK3R2 , OGFRL1 , and NSA2 ) exhibiting differential expression in opposite directions ( Fig 4B; S17 Table ) . Six genes were common between the −5D/10-days and −5D/5-weeks groups , with all genes expressed in the same direction in both groups ( Fig 4D; S18 Table ) . Three genes were common between the −5D/10-days and +5D/10-days groups , all expressed in the same direction ( Fig 5B; S19 Table ) , indicating that they were sensitive to defocus but not to the sign of defocus . However , 13 out of 18 genes , which were common between the −5D/5-weeks and +5D/5-weeks groups , exhibited sign-of-defocus–specific expression ( Fig 5D; S20 Table ) . These included nine coding genes ( ZC3H11A , TRIM23 , STARD3NL , RCBTB1 , PPP2CA , LOC100394842 , CUL3 , COMMD3 , ACTR8 ) and four long noncoding RNAs ( LOC103794697 , LOC100396694 , LOC100394543 , LOC100392587 ) . Thus , these data reveal that hyperopic and myopic defocus affect the expression of largely different genes in the retina . Very few genes are affected by both hyperopic and myopic defocus and change direction of expression in response to the defocus of different sign . GO analysis showed that the early ( 10 days ) and sustained ( 5 weeks ) response to defocus involved largely the same cellular functions ( albeit to a different extent ) ( Fig 4E and 4F; S9–S12 Tables ) ; however , there were several cellular functions unique to the sustained response to defocus ( Fig 4E and 4F; S9–S12 Tables ) . Specifically , the response to 5 weeks of myopic defocus imposed by +5D lenses ( associated with reduced axial eye growth ) involved changes in metabolism , cell signaling , protein trafficking , and posttranslational modification of proteins ( Fig 4E; S11 and S12 Tables ) , whereas the sustained response to hyperopic defocus imposed by −5D lenses ( associated with increased axial eye growth ) involved substantial increase in gene expression , protein trafficking , and vitamin and mineral metabolism ( Fig 4F; S9 and S10 Tables ) . Moreover , pathway analysis revealed almost complete transition from one set of pathways to another between 10 days of treatment and 5 weeks of treatment for both −5D and +5D experimental groups ( Fig 4G and 4H; S13–S16 Tables ) . For myopic defocus imposed by positive lenses , the most prominent change is a transition from RANK , SAPK/JNK , NGF , and gap junction signaling to protein ubiquitination pathway , Notch signaling , glutamate receptor signaling , oncostatin M signaling , α-adrenergic signaling , and ephrin B signaling ( Fig 4G; S15 and S16 Tables ) . For hyperopic defocus imposed by negative lenses , there is an apparent transition from ephrin A and reelin signaling to ß-adrenergic signaling , Gα12/13 signaling , calcium signaling , glutamate receptor signaling , TNFR1 signaling , HIPPO signaling , RAN signaling , NOS signaling , synaptic long-term potentiation , relaxin signaling , and dopamine receptor signaling ( Fig 4H; S13 and S14 Tables ) . In summary , there is almost complete transition from one set of pathways to another over time when the retina responds to sustained optical defocus . Comparison of the cellular functions affected by different defocus in the −5D and +5D experimental groups revealed that although most cellular functions affected by imposed hyperopic or myopic defocus were the same ( Fig 5E and 5F; S9–S12 Tables ) , there was a clear increase in gene expression during the first 10 days after exposure to myopic defocus imposed by +5D lenses and a substantial increase in RNA trafficking and RNA posttranscriptional modifications after exposure to +5D lenses in both 10-days and 5-weeks groups ( Fig 5E and 5F; S9–S12 Tables ) , indicating that alternative splicing and isoform switching play important roles in the eye’s response to positive lenses . Furthermore , pathway analysis revealed that positive and negative lenses influenced different sets of pathways ( Fig 5G and 5H; S13–S16 Tables ) . During the first 10 days , the retina responded to hyperopic defocus imposed by −5D lenses with changes in glycogen and S-methyl-5′-thioadenosine degradation , spermine and choline biosynthesis , and ephrin receptor signaling , whereas myopic defocus imposed by +5D lenses caused changes in phenylalanine degradation , RANK , NGF , and gap junction signaling ( Fig 5G; S13 and S15 Tables ) . Five weeks of exposure to optical defocus resulted in large-scale changes in retinal signaling . The −5D-lens-imposed hyperopic defocus primarily affected ß-adrenergic and cAMP-mediated signaling , tRNA splicing , protein kinase A signaling , Gα12/13 signaling , calcium signaling , dopamine-DARPP32 feedback signaling , G-protein coupled receptor signaling , glutamine biosynthesis , nNOS signaling , RAN , HIPPO , and TNFR1 signaling as well as the Wnt/Ca+ pathway , among others ( Fig 5H; S14 and S16 Tables ) . Retinal response to 5 weeks of myopic defocus imposed by +5D lenses involved protein translation and protein ubiquitination pathways , oncostatin M and somatostatin receptor 2 signaling , CNTF signaling , aldosterone signaling , JAK/Stat signaling , ephrin B signaling , integrin signaling , interleukin signaling , Notch signaling , α-adrenergic signaling , glucocorticoid receptor signaling , and mTOR signaling , among several other pathways ( Fig 5H; S14 and S16 Tables ) . Taken together , these data suggest that postnatal eye growth and refractive development are regulated by a bidirectional mechanism that involves active stimulation of eye growth by hyperopic defocus and active suppression of eye growth by myopic defocus , through largely independent pathways . Genes comprising signaling pathways that underlie physiological processes are often targeted by mutations , causing human diseases . To identify candidate genes involved in myopia development in humans , we compared the genes differentially expressed in the marmoset retina in response to imposed defocus with a list of genes located within QTLs found to be associated with myopia in the human population ( S21 Table ) [47] . This analysis revealed that a total of 29 differential genes identified in this study were localized within 24 human myopia QTLs ( Fig 6; S22–S24 Tables ) , including two genes differentially expressed in the −5D/10-days group ( Fig 6A and 6C ) , nine genes differentially expressed in the −5D/5-weeks group ( Fig 6A and 6C ) , and 21 genes differentially expressed in the +5D/5-weeks group ( Fig 6B and 6C ) . The overlap between myopia candidate genes and genes differentially expressed in the −5D/5-weeks and +5D/5-weeks groups was statistically significant ( OR = 4 . 2 , P < 4 . 9 × 10−4; OR = 3 . 9 , P < 7 . 5 × 10−7 , respectively ) , indicating functional connection between genes found by the genetic mapping studies in humans and genes differentially expressed in the marmoset retina upon exposure to optical defocus . These data also suggest that approximately 24% of the human QTLs may be associated with genetic variations in the genes identified in marmosets as regulating retinal response to defocus .
While there is plentiful evidence that eye growth can be modulated by imposed optical defocus of different signs , there remains little consensus regarding the mechanism underlying the eye’s sensitivity to the sign of optical defocus . Nevin and colleagues [24] reported that preventing access to sharp vision prevented refractive compensation for the sign of defocus , arguing that eye growth responds to blur and not the sign of defocus . Conversely , other findings support the alternative view that the eye can detect the sign of defocus and adjust its growth accordingly to compensate for either hyperopic or myopic defocus even when imposed on an image that is significantly blurred , indirectly supporting the BESOD model of emmetropization [48–50] . Furthermore , the sign-of-defocus–sensitive expression of the retinal transcription factor ZENK/Egr1 led to the hypothesis that both myopic and hyperopic defocus would affect expression of the same genes in the retina but in opposite direction , presumably activating or suppressing a single pathway controlling eye growth [51] . Our data provide direct evidence that the retina detects myopic and hyperopic defocus separately through largely distinct and independent pathways , supporting the existence of BESOD ( Fig 7 ) . Furthermore , we found that there is very little overlap between the genes and pathways underlying response to the myopic defocus and the genes and pathways underlying response to the hyperopic defocus . We found only 13 genes that exhibited sign-of-defocus–sensitive expression , i . e . , ACTR8 , COMMD3 , CUL3 , LOC100392587 , LOC100394543 , LOC100394842 , LOC100396694 , LOC103794697 , PPP2CA , RCBTB1 , STARD3NL , TRIM23 , and ZC3H11A . These genes represent about 1% of all genes affected by the optical defocus in the retina , and we speculate that they may act as switches that trigger largely distinct sign-of-defocus–specific signaling cascades underlying BESOD . Our findings also support the existence of a switch from one set of genes to another when the retina transitions from an early response to defocus to a later , possibly sustained , response . There was almost complete transition from one set of pathways at 10 days to different pathways after 5 weeks of treatment with both myopic and hyperopic defocus imposed by the positive and negative lenses , respectively . The transition from the early response to the later response in the retinae exposed to myopic defocus was associated with significant changes in cell signaling , protein synthesis and degradation , posttranslational modification of proteins , protein trafficking , and energy production . This was accompanied by the transition from the pathways regulating phenylalanine degradation , RANK , SAPK/JNK , NGF , and gap junction signaling to the pathways that regulate ephrin B , α-adrenergic , iNOS , fatty acid α-oxidation , oncostatin M , glutamate receptor , Notch , and protein ubiquitination signaling , among other pathways . In the retinae exposed to hyperopic defocus , transition from the early response to the later response was associated with changes in gene expression , protein trafficking , and vitamin and mineral metabolism . It was accompanied by the transition from the pathways regulating glycogen degradation , spermine and spermidine biosynthesis , ephrin A and reelin signaling to the pathways that control synaptic long-term potentiation and phototransduction as well as dopamine receptor , TNFR1 , HIPPO , RAN , nNOS , glutamate receptor , calcium , and ß-adrenergic signaling , among other pathways . We also found that although response to negative lenses appeared to be stronger compared to positive lenses during the first 10 days of exposure , positive lenses elicited a substantially stronger response at the level of gene expression after 5 weeks of treatment . This suggests that the time course of retinal response to hyperopic and myopic defocus might be slightly different and that positive lenses ultimately may have greater impact on eye growth than negative lenses . Our data also suggest that positive and negative lenses affect different sets of genes and pathways . The most significant differences between positive and negative lenses after 10 days of treatment were related to the cellular functions underlying RNA trafficking and RNA splicing , which may indicate that initial exposure to defocus activates a large-scale switch from one set of protein isoforms to another . Moreover , negative lenses primarily affected glycogen degradation , spermine and spermidine biosynthesis , and ephrin A and reelin signaling , whereas positive lenses induced changes in gap junction , NGF , and RANK signaling as well as phenylalanine degradation . Positive and negative lenses also affected significantly different cellular functions and pathways after 5 weeks of treatment . The largest difference between the 5-weeks groups were significant changes in RNA trafficking and splicing induced by positive lenses , which again suggests a large-scale switch from one set of protein isoforms to another . While negative lenses primarily affected pathways regulating ß-adrenergic signaling , cAMP-mediated signaling , calcium signaling , dopamine signaling , G-protein coupled receptor signaling , glutamine biosynthesis , nNOS signaling , RAN signaling , HIPPO signaling , TNFR1 signaling , Wnt/Ca+ pathway , and synaptic long-term potentiation , positive lenses caused changes in pathways regulating protein ubiquitination , fatty acid α-oxidation , and tryptophan degradation as well as EIF2 , oncostatin M , CNTF , aldosterone , JAK/Stat , ephrin B , integrin , Notch , glucocorticoid receptor , mTOR , iNOS , and α-adrenergic signaling , among other pathways . Although many pathways that we found to underlie the retinal responses to defocus are novel , there are several pathways that had previously been implicated in the development of myopia . For example , a mutation in the gene NDUFAF7 , which encodes a mitochondrial reduced nicotinamide adenine dinucleotide ( NADH ) dehydrogenase complex assembly factor , was associated with pathological myopia in humans , implicating mitochondrial dysfunction in refractive error development [55] . Protein ubiquitination may be associated with myopia development , because a loss-of-function mutation in the E3 ubiquitin-protein ligase UBE3B was shown to cause Kaufman oculocerebrofacial syndrome , which includes myopia as one its prominent features [56] . Two genome-wide association studies found a strong association between refractive error and GJD2 , which encodes gap junction protein 2 expressed in the retina , suggesting that gap junction signaling may be involved in refractive eye development [57 , 58] . Glutamate receptor signaling was found to be involved in myopia development by two genetic linkage studies [58 , 59] and also in a study of a glutamate receptor Grik2 mouse knockout model [36] . Aspartate receptors were implicated in myopia development by at least two studies [60 , 61] . Nitric oxide and cAMP signaling were also suggested to play a role in the development of myopia by several studies in chickens [37–39] . Dopamine signaling has long been associated with experimental myopia and refractive eye development [40] . Riddell and colleagues also found that pathways regulating apoptosis , oxidative phosphorylation , and mesenchymal development are involved in refractive development in chickens and humans [41 , 43] . Our finding that myopic and hyperopic defocus signals driving eye growth in opposite directions propagated via largely different retinal pathways is supported by several functional studies , which found that pharmacological agents that inhibit negative-lens-induced myopia are completely different from those inhibiting positive-lens-induced hyperopia [62–64] . Moreover , the temporal features of flickering light that suppress refractive error development are different for imposed myopia and hyperopia [65] . We also note that many of the genes changing expression in response to retinal defocus in this study are localized within human QTLs linked to myopia . This strongly suggests that there is a functional overlap between the genes that regulate retinal response to defocus and the genes that cause human myopia . What creates conditions for the development of myopia and why myopia progresses in some individuals and not in others remain important topics of further research . Several treatment options for myopia , such as optical correction using spectacles , multifocal contact lenses , or atropine , are available , but efficacy is limited [66–72] . The identification of drug targets and the development of suitable drugs that can be used to suppress or possibly even reverse the progression of myopia require a deeper understanding of the signaling pathways underlying the ocular response to optical defocus of different signs . The results of this study show that the retina can distinguish between myopic and hyperopic defocus and responds to defocus of opposite signs by activating largely distinct pathways . Identification of these pathways provides a framework for understanding the molecular mechanisms underlying the visual control of emmetropization and the identification of new drug targets for the development of more effective treatment options for myopia .
Marmosets were bred and raised in the animal care facilities at the New England College of Optometry and at the State University of New York , College of Optometry , according to the United States Department of Agriculture ( USDA ) standards for animal care and use , and fully complied with the guidelines outlined in the Weatherall report on the use of nonhuman primates in research . All procedures also adhered to the Association for Research in Vision and Ophthalmology ( ARVO ) statement for the use of animals in ophthalmic and vision research and were approved by the New England College of Optometry and SUNY College of Optometry Institutional Animal Care and Use Committees ( Protocol Nos . DT-1 . 10 . 07 and DT-2011-06-1 , respectively ) . Animals were anesthetized via intramuscular injection of alfaxalone ( 1 . 5 mg/kg ) and were humanely killed by intracardiac pentobarbital ( 100 mg/kg ) overdose while under deep surgical anesthesia . Twelve common marmosets ( C . jacchus ) were used in this study . Marmosets were housed under lighting provided from daylight balanced fluorescent lamps on a 10-hour:14-hour light-dark cycle . Temperature was maintained at 75 ± 2°F with 45% ± 5% humidity . Food and water were provided ad libitum within the home cage and consisted of a formulated dry pellet ( Mazuri New World Diet , Mazuri Exotic Animal Nutrition , St . Louis , MO ) with regularly varied supplements of fresh fruit and protein . Custom-made soft single vision contact lenses were used to impose defocus . Animals of both sexes , which were 74 ± 5 days old at the beginning of the experiment , were fitted with either a +5D ( imposed myopic defocus ) or −5D ( imposed hyperopic defocus ) contact lens on one eye for either 10 days or 5 weeks . The contralateral eye wore a plano lens and was used as an interocular control . All lenses were inserted at the beginning of the light period and removed daily at the beginning of the dark period . Changes in axial eye growth and refractive state were measured at the onset and at the end of lens treatment . Refractive state was measured by retinoscopy and Hartinger coincidence refractometry following cycloplegia with 1% cyclopentolate . Refractive state was taken as the average of the spherical equivalents from both measures . Axial length was measured with A-scan ultrasonography and reported as changes in vitreous chamber depth ( on-axis distance from the posterior surface of the crystalline lens to the inner surface of the retina ) . Our goal was to explore gene expression during detection of and active compensation for imposed defocus , when biometric changes cannot be readily detected . Our selection of the 10-day and 5-week rearing durations was guided by our previously reported data from two independent groups of age-matched marmosets ( eight animals each ) , which were raised exposed to either imposed myopic defocus ( with +5D lenses ) or imposed hyperopic defocus ( with −5D lenses ) until they achieved full compensation . In these animals , we obtained the time course of changes in refractive state and depth of vitreous chamber in response to the imposed myopic and hyperopic defocus . See Ref . [25] for details . After lens treatments , animals were humanely killed following IACUC-approved protocols . Both lens-treated and control eyes were enucleated , the retinae were dissected from the enucleated eyes , and the choroid/RPE was removed . The retinae were washed in RNAlater ( Thermo Fisher Scientific , Grand Island , NY ) for 5 minutes , frozen in liquid nitrogen , and stored at −80°C until processed for this study . To isolate RNA , tissue samples were homogenized at 4°C in a lysis buffer using Bead Ruptor 24 tissue homogenizer ( Omni , Kennesaw , GA ) . Total RNA was extracted from each tissue sample using miRNAeasy mini kit ( QIAGEN , Germantown , MD ) following the manufacturer’s protocol . The integrity of RNA was confirmed by analyzing 260/280 nm ratios ( Ratio260/280 = 2 . 11–2 . 13 ) on a Nanodrop ( Thermo Fisher Scientific , Grand Island , NY ) and the RNA Integrity Number ( RIN = 9 . 0–10 . 0 ) using Agilent Bioanalyzer ( Agilent , Santa Clara , CA ) . Illumina sequencing libraries were constructed from 1 μg of total RNA using the TruSeq Stranded Total RNA LT kit with the Ribo-Zero Gold ribosomal RNA depletion module ( Illumina , San Diego , CA ) . The libraries , each containing a specific index ( barcode ) , were pooled at equal concentrations using the randomized complete block ( RCB ) experimental design before sequencing on Illumina HiSeq 2500 sequencing system ( Illumina , San Diego , CA ) . The number of libraries per multiplexed sample was adjusted to ensure sequencing depth of about 70 million reads per library ( paired-end , 2 × 100 nucleotides ) . The actual sequencing depth was 70 , 851 , 538 ± 10 , 430 , 968 with read quality score 39 . 7 ± 0 . 2 . The FASTQ raw data files generated by the Illumina sequencing system were imported into Partek Flow software package ( version 6 . 0 . 17 . 0723 , Partek , St . Louis , MO ) , libraries were separated based on their barcodes , adapters were trimmed , and remaining sequences were subjected to pre-alignment quality control using Partek Flow pre-alignment QA/QC module . After the assessment of various quality metrics , bases with the quality score <37 were removed ( ≤5 bases ) from each read . Sequencing reads were then mapped to the marmoset reference genome C . jacchus 3 . 2 . 1 ( NCBI ) using the STAR aligner ( version 2 . 5 . 2b ) , resulting in 85 . 9% ± 2 . 4% mapped reads per library , which covered 36 . 9% ± 1 . 8% of the genome . Aligned reads were quantified to transcriptome using Partek E/M annotation model and the NCBI C . jacchus 3 . 2 . 1 annotation GFF file to determine read counts per gene/genomic region . The generated read counts were normalized by the total read count and subjected to the Partek Flow Gene Specific Analysis ( GSA ) to detect differentially expressed transcripts . Data were simultaneously fitted with Normal , Lognormal , Lognormal with shrinkage , Negative Binomial , or Poisson statistical models , and the best model was then applied to the corresponding subset of transcripts depending on the transcript expression within each subset . Differentially expressed transcripts were identified using a P value threshold of 0 . 05 adjusted for genome-wide statistical significance using Storey’s q-value algorithm [73] . Differential expression was calculated as fold change ( FC , lens-treated eye versus control ) . To identify sets of genes with coordinate expression , differentially expressed transcripts were clustered using Partek Flow hierarchical clustering module using average linkage for the cluster distance metric and Euclidean distance metric to determine the distance between data points . Each RNA-seq sample corresponding to one control or one lens-treated eye was analyzed as a biological replicate , thus resulting in three biological replicates for control eyes and three biological replicates for lens-treated eyes in each of the four experimental groups . To identify biological functions ( GO categories ) , which are significantly affected by the changes in gene expression induced by the optical defocus in the retina , we used QIAGEN’s Ingenuity Pathway Analysis ( IPA ) software and database ( QIAGEN , Germantown , MD ) . IPA Downstream Effects Analysis module was used to visualize biological trends and predict the effect of gene expression changes in the datasets on biological processes . Downstream Effects Analysis is based on expected causal effects between genes and functions; the expected causal effects are derived from the literature compiled in the Ingenuity Knowledge Base . The analysis examines genes in the dataset that are known to affect functions , compares the genes’ direction of change to expectations derived from the literature , and then issues a prediction for each function based on the direction of change . The direction of change was determined as the difference in gene expression in the retina exposed to +5D or −5D lenses relative to the contralateral control retina exposed to the plano lens . We used the Fisher's exact test with a P value threshold of 0 . 05 to estimate statistical significance and the z-score algorithm to make predictions about direction of change . The z-score algorithm is designed to reduce the chance that random data will generate significant predictions . The activation z‐score was used to infer likely activation states of biological functions based on comparison with a model that assigns random regulation directions . The z-score provided an estimate of statistical quantity of change for each biological function found to be statistically significantly affected by the changes in gene expression . The activation z‐score can also predict implicated biological functions independently from their associated P value , based on significant pattern match of up- or down-regulation . The activation z-score was also employed in the IPA Pathways Activity Analysis module to predict activation or suppression of the canonical pathways affected by optical defocus . The significance values for the canonical pathways were calculated by the right-tailed Fisher's exact test . The significance indicates the probability of association of molecules from each dataset with the canonical pathway by random chance alone . Pathways Activity Analysis determined if canonical pathways , including functional end points , are activated or suppressed based on differentially expressed genes or proteins in each dataset . Once statistically significant biological functions and canonical pathways for −5D/10 days , −5D/5 weeks , +5D/10 days , and +5D/5 weeks datasets were identified , we subjected these datasets to the Core Functional Analysis in IPA to compare the datasets and identify key similarities and differences in the biological functions and canonical pathways affected in each experimental group . To identify candidate genes for human myopia , we compared the genes differentially expressed in the marmoset retina in response to imposed defocus with a list of genes located within QTLs found to be associated with myopia in a human population . We first compiled a list of all SNPs or markers exhibiting statistically significant association with myopia in the human linkage or GWAS studies . LDlink’s LDmatrix tool ( National Cancer Institute ) was used to identify SNPs in linkage disequilibrium and identify overlapping chromosomal loci . We then used UCSC Table Browser to extract all genes located within critical chromosomal regions identified by the human linkage studies or within 200 kb ( ±200 kb ) of the SNPs found by GWAS . The list of genes located within human QTLs was compared with the list of genes that we found to be differentially expressed in the marmosets exposed to optical defocus using Partek Genomics Suite . The statistical significance of the overlaps was estimated using probabilities associated with the hypergeometric distribution , using Bioconductor software package GeneOverlap and associated functions .
|
The worldwide prevalence of myopia is predicted to increase from the current 23% to about 50% in the next three decades . Although much effort has been directed towards elucidating the mechanisms underlying refractive eye development and myopia , treatment options for myopia are mostly limited to optical correction , which does not prevent progression of myopia nor the pathological blinding complications often associated with the disease . Several experimental optics-based treatments have had only limited effect on myopia progression , and currently available drug treatments are limited and the mechanisms of action are not well understood . The development of safe and effective pharmacological treatments for myopia is urgently needed to prevent the impending myopia epidemic . The main obstacles that prevent the development of anti-myopia drugs are the uncertainties regarding the mechanisms controlling eye growth and optical development , including the molecular signaling pathways underlying it . In this study , we show that , contrary to the conventional thinking that myopic and hyperopic defocus trigger opposite changes in the same genes and pathways to guide postnatal eye growth , defocus of opposite signs affect eye growth via largely distinct retinal pathways . Knowing that myopic and hyperopic defocus signals drive eye growth in opposite directions and propagate via different pathways provides a framework for the development of new anti-myopia drugs . Myopia can be controlled pharmacologically by stimulating pathways underlying the retinal response to positive lenses and/or by suppressing pathways underlying the retinal response to negative lenses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"eye",
"lens",
"medicine",
"and",
"health",
"sciences",
"methods",
"and",
"resources",
"ocular",
"anatomy",
"marmosets",
"vertebrates",
"animals",
"mammals",
"notch",
"signaling",
"animal",
"models",
"primates",
"cell",
"signaling",
"experimental",
"organism",
"systems",
"eyes",
"calcium",
"signaling",
"research",
"and",
"analysis",
"methods",
"monkeys",
"animal",
"studies",
"new",
"world",
"monkeys",
"gene",
"expression",
"myopia",
"head",
"signal",
"transduction",
"visual",
"impairments",
"retina",
"eukaryota",
"anatomy",
"cell",
"biology",
"ophthalmology",
"genetics",
"biology",
"and",
"life",
"sciences",
"ocular",
"system",
"amniotes",
"organisms"
] |
2018
|
Gene expression in response to optical defocus of opposite signs reveals bidirectional mechanism of visually guided eye growth
|
Crusted scabies , or hyperinfestation with Sarcoptes scabiei , occurs in people with an inadequate immune response to the mite . In recent decades , data have emerged suggesting that treatment of crusted scabies with oral ivermectin combined with topical agents leads to lower mortality , but there are no generally accepted tools for describing disease severity . Here , we describe a clinical grading scale for crusted scabies and its utility in real world practice . In 2002 , Royal Darwin Hospital ( RDH ) , a hospital in tropical Australia developed and began using a clinical grading scale to guide the treatment of crusted scabies . We conducted a retrospective observational study including all episodes of admission to RDH for crusted scabies during the period October 2002–December 2010 inclusive . Patients who were managed according to the grading scale were compared with those in whom the scale was not used at the time of admission but was calculated retrospectively . There were 49 admissions in 30 patients during the study period , of which 49 ( 100% ) were in Indigenous Australians , 29 ( 59% ) were male and the median age was 44 . 1 years . According to the grading scale , 8 ( 16% ) episodes were mild , 24 ( 49% ) were moderate , and 17 ( 35% ) were severe . Readmission within the study period was significantly more likely with increasing disease severity , with an odds ratio ( 95% CI ) of 12 . 8 ( 1 . 3–130 ) for severe disease compared with mild . The patients managed according to the grading scale ( 29 episodes ) did not differ from those who were not ( 20 episodes ) , but they received fewer doses of ivermectin and had a shorter length of stay ( 11 vs . 16 days , p = 0 . 02 ) . Despite this the outcomes were no different , with no deaths in either group and a similar readmission rate . Our grading scale is a useful tool for the assessment and management of crusted scabies .
Scabies is a parasitic infestation caused by the mite Sacroptes scabiei var hominis . Globally , over 300 million people are estimated to be affected [1] . The mite is endemic in disadvantaged and impoverished communities [2] , [3] . In Australia Indigenous people suffer a significant disadvantage in health outcomes compared with non-Indigenous Australians [4] , [5] , and scabies is endemic in many Indigenous communities in northern Australia , with a recent survey demonstrating a mean prevalence of 13 . 4% in five remote Indigenous communities [6] . Crusted scabies ( also known as “Norwegian scabies” ) is hyperinfestation with the Sarcoptes scabiei mite , and is characterized by a non-protective host immune response , the development of hyperkeratotic skin crusts and skin fissuring [7] . It is a severe disease with a significantly higher mortality than ordinary scabies . Unlike ordinary scabies , where there are usually less than 20 mites on the host's entire skin , individuals with crusted scabies can have up to 4000 mites per gram of skin and are extremely infectious to others [8] , [9] . Despite the severity of the disease there is significant variability in the clinical presentation , and there is currently no generally accepted method of describing the severity of a crusted scabies infection . The optimal treatment for crusted scabies has not been subjected to a comparative trial and is generally based on expert opinion [10] , [11] . However observational data suggest that the use of multiple doses of oral ivermectin as therapy for crusted scabies can lead to a significant decline in mortality [9] , [12] , [13] . In an attempt to formalize and improve the treatment of crusted scabies , we developed a grading scale , based on our clinical experience in managing such patients . This was introduced into routine clinical use at our hospital in 2002 , and has been used since this time to titrate the duration of ivermectin and topical therapy to illness severity . Here , we describe the grading scale and our experience with it over the first eight years of its use . We aimed to evaluate the utility of the grading scale , including its correlation with other putative markers of illness severity , the safety of its use and the effect on length of stay and relapse rates .
The study was approved by the human research ethics committee of the Menzies School of Health Research and Northern Territory Department of Health . 350 bed tertiary referral hospital in the tropical Northern Territory , Australia , serving a population of approximately 150 , 000 people spread over an area of 500 , 000 km2 , including many remote Indigenous communities . Local policies encourage the hospitalization of patients with crusted scabies for clinical management , as well as environmental health input to address the risk of ongoing transmission in an index patient's household . The standard treatment protocol for crusted scabies includes prolonged hospitalization in a single room with contact precautions , the use of topical benzyl benzoate plus 5% tea tree oil 2–3 times per week [14] , multiple doses of oral ivermectin ( as described below ) , topical keratolytics , systemic antibacterial drugs where judged clinically necessary , and attention to medical comorbidities . All patients admitted to our hospital with a discharge diagnosis of crusted scabies between 1st of October 2002 and 31st of December 2010 were included in the study . Crusted scabies was diagnosed based on the clinical opinion of an Infectious Diseases specialist , supplemented by skin scrapings demonstrating S . scabiei mites on microscopy . The grading scale for crusted scabies is shown in Figure 1 . It is based on clinical assessment in four key areas: the distribution and extent of crusting; the depth of crusting; the degree of skin cracking and pyoderma; and the number of previous episodes . This scale was developed in 2002 by two of the authors ( JD and BC ) for use with all patients hospitalised with crusted scabies . It was partly based on previous local experience that multiple doses of ivermectin in addition to topical treatment were more effective than topical treatment alone for the treatment of crusted scabies [9] . Other studies have confirmed the efficacy of the combination of ivermectin and topical therapy for crusted scabies [13] , [15] , [16] . During the study period medical staff managing patients with crusted scabies were encouraged but not compelled to use the grading scale to guide management . Therefore we were able to compare those patients in whom the grading scale was applied at the time of the patient's clinical presentation to those in whom the grading scale was not used and then calculated retrospectively by the authors . We reviewed clinical notes , bedside charts and the hospital's clinical pathology database for each patient using a standardized case record form . We collected data on demographics , comorbidities , disease severity , grading scale and outcomes . Where the grading scale had not been prospectively documented , we calculated it based on the detailed clinical information found in the medical record . Each admission ( rather than each individual patient ) was counted as a discrete episode . Where a patient had more than one admission during the study period , it was only counted as a separate episode if at least 30 days had elapsed from the previous date of discharge . Iatrogenic immunosuppresion was defined as the use of any of the following medications within the past 3 months: prednisolone ≥0 . 5 mg/kg/day or equivalent for at least 14 days; immunosuppresion for solid organ transplant; cancer chemotherapy; immunosuppressive monoclonal antibody use; any other use of azathioprine , methotrexate , leflunomide , cyclosporine , mycophenolate , or cyclophosphamide . Hazardous alcohol use was defined as an average of >4 standard drinks per day for a man or >2 for a woman . Chronic renal disease was defined as an estimated glomerular filtration rate of less than 30 ml/min , or the need for dialysis . Data were entered into a purpose-built database using Epidata v 3 . 0 and were analysed using Stata version 10 ( Statacorp , College Station , Texas , USA ) . Categorical variables were compared using Fisher's exact test , and continuous using Mann-Whitney-U test . Correlations were assessed using Spearman's rank correlation . P values of <0 . 05 were considered significant
There were 49 admissions for crusted scabies in 30 patients during the eight year study period . Of the episodes , 49 ( 100% ) were in Indigenous Australians , 29 ( 59% ) were male and the median age at the time of the first admission within the study period was 45 . 4 years ( Table 1 ) . Most of the patients lived in remote areas , and iatrogenic immunosuppresion was rare . All patients received at least one dose of oral ivermectin ( with a mean of 5 . 2 doses , and a range of 2 to 10 ) . 47 patients ( 95% ) were treated with topical benzyl-benzoate in combination with 5% tea-tree oil , and the remainder with topical permethrin . In addition , all patients were treated with topical Calmurid ( lactic acid and urea in sorbolene cream , used as a keratolytic ) . Systemic antibiotics were used in 38 ( 79% ) of episodes . According to the grading scale , 8 ( 16% ) episodes were mild ( grade 1 ) , 24 ( 49% ) were moderate ( grade 2 ) , and 17 ( 35% ) were severe ( grade 3 ) . Seven episodes ( 14% ) were complicated by bacteraemia , with the causative organism being Staphylococcus aureus in 6 patients , and a mixed infection with Group A streptococcus and Escherischia coli in 1 . The disease severity according to the grading scale did not correlate with the proportion of patients with bacteremia , or with the peak plasma C-reactive protein during the admission ( table 2 ) . However , there was a non-significant trend towards lower nadir plasma albumin and longer hospital stay with higher severity ( table 2 ) . No patients in this cohort died during the hospital admission , but a substantial proportion ( 47% ) required readmission for crusted scabies within the eight year study period . Readmission was significantly more likely with increasing disease severity , with an odds ratio ( 95% CI ) of 5 . 9 ( 0 . 7–55 . 9 ) for moderate disease compared with mild , and 12 . 8 ( 1 . 3–130 ) for severe disease compared with mild . There was no significant difference in age , gender , location of residence or comorbidities between those patients who had the severity score calculated at the time of admission ( n = 29 ) and those who did not ( n = 20 ) . Episodes where the grading scale was calculated at the time of admission had a significantly shorter length of stay , and received fewer doses of ivermectin than those not managed using the grading scale ( Table 3 ) . Despite this their outcomes were no different , with no deaths in either group , and a similar readmission rate in the two groups .
This is the first published description of a clinical severity grading scale for use in patients with crusted scabies . The use of this grading scale in our setting is associated with good outcomes despite shorter hospital stays and less ivermectin use compared with those managed without the use of the grading scale . Crusted scabies is a severe disease with significant morbidity and mortality which is more prevalent in communities such as remote-dwelling Australian Indigenous people [2] . Crusted scabies is usually reported as occurring in patients who are immunosuppressed , either iatrogenically [17] , [18] , [19] , [20] or by retroviral infection [21] , [22] , [23] . In our cohort there was a high rate of hazardous alcohol use , diabetes and chronic renal disease , but only 16% of episodes were associated with iatrogenic immunosuppresion or HTLV-1 infection . This reinforces the findings of previous studies that , in Indigenous Australians , the majority of people with crusted scabies do not meet the generally accepted definitions of significant immunosuppression and suggests that the immune defect in patients with crusted scabies is subtle and probably multifactorial [8] . Our grading scale did not correlate with many of the putative measures of disease severity we used ( CRP , ICU admission , bacteraemia ) . However , these factors are really measures of the sequelae of crusted scabies and there is no generally accepted single marker of disease severity in this setting ( hence the need for the clinical grading scale ) . The degree of systemic inflammation and risk of bacteraemia are likely to relate to multiple factors , including the patient's immune responses , the depth of skin cracks , the degree of bacterial skin colonization and the patient's underlying comorbidities . Hence this lack of correlation does not necessarily imply that the grading scale does not reflect disease severity . Long hospital stays ( particularly those involving single rooms and contact isolation ) are expensive to the health care system , and frustrating for patients . The 5 day decrease in length of stay which we observed with the use of the grading scale , with no increase in relapse rates , is substantial and represents a large cost saving . Another potential advantage of our grading scale is that it may help guide the duration and type of therapy for those clinicians who are less experienced in the management of crusted scabies . Given that crusted scabies is a rare condition in most settings , the utility of such a grading scale for the average clinician is a good reason for its use . Ivermectin is an orally administered semi-synthetic macrocyclic lactone antibiotic . It is approved for the treatment of scabies in France but is not licensed for the treatment of scabies in the United States , United Kingdom or Australia . However is it commonly used off-label for the treatment of scabies in Australia . Ivermectin does not sterilize scabies eggs so multiple doses are recommended to kill newly hatched mites [10] . Ivermectin has been associated with adverse effects in some studies , which emphasizes the benefit of using a grading scale that allows for the titration of the total dose of ivermectin and in our study a possible reduction in number of doses in patients with milder disease . Intensive ivermectin use may also increase the probability of the mite developing resistance especially in patients with multiple relapses [24] . This study was planned prospectively , but the grading scale had to be calculated retrospectively in 40% of patients , introducing possible inaccuracies in the calculated scores . Fortunately , a detailed clinical assessment was recorded in the medical record for all patients , and thus we were able to calculate the score for all patients without having to interpolate missing data . Despite this , the score calculated at the time of clinical assessment is likely to be more accurate; retrospectively calculated scores may have underestimated severity in certain areas such as degree of crusting and shedding . However , this would not affect the overall conclusions regarding the use of the score , as the outcome measures were not the scores themselves , but objective measures including length of hospital stay and need for re-admission . Our population differs substantially from some others in whom crusted scabies has been reported to occur . Hence it is important for the grading scale to be studied in other populations before concluding that it is useful in all settings . We have described a simple clinical grading scale to aid in the management of patients with crusted scabies . If validated in other settings , its use is likely to improve the management of crusted scabies and may lead to a decreased length of required hospital stay and of ivermectin treatment , without compromising outcomes .
|
Crusted scabies is a severe skin condition caused by a microscopic parasitic mite . It occurs in people whose immune system does not react properly to the mite and it leads to crusting and cracking of the skin and can cause death . The usual treatment for crusted scabies is a tablet called ivermectin combined with anti-scabies skin creams . However , there is no current method of measuring the severity of crusted scabies and thus deciding how long to continue the treatment for . We have developed a grading scale based on examination of the skin , which classifies patients as mild , moderate or severe , and uses this grading to suggest the duration of treatment . We have trialed this grading scale over an 8-year period in 49 episodes of crusted scabies requiring hospital admission , and have found that it leads to a shorter length of hospital stay and treatment , but equivalent outcomes compared to those who were treated without the use of the grading scale .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2013
|
A Novel Clinical Grading Scale to Guide the Management of Crusted Scabies
|
Refractive error is a highly heritable quantitative trait responsible for considerable morbidity . Following an initial genome-wide linkage study using microsatellite markers , we confirmed evidence for linkage to chromosome 3q26 and then conducted fine-scale association mapping using high-resolution linkage disequilibrium unit ( LDU ) maps . We used a preliminary discovery marker set across the 30-Mb region with an average SNP density of 1 SNP/15 kb ( Map 1 ) . Map 1 was divided into 51 LDU windows and additional SNPs were genotyped for six regions ( Map 2 ) that showed preliminary evidence of multi-marker association using composite likelihood . A total of 575 cases and controls selected from the tails of the trait distribution were genotyped for the discovery sample . Malecot model estimates indicate three loci with putative common functional variants centred on MFN1 ( 180 , 566 kb; 95% confidence interval 180 , 505–180 , 655 kb ) , approximately 156 kb upstream from alternate-splicing SOX2OT ( 182 , 595 kb; 95% CI 182 , 533–182 , 688 kb ) and PSARL ( 184 , 386 kb; 95% CI 184 , 356–184 , 411 kb ) , with the loci showing modest to strong evidence of association for the Map 2 discovery samples ( p<10−7 , p<10−10 , and p = 0 . 01 , respectively ) . Using an unselected independent sample of 1 , 430 individuals , results replicated for the MFN1 ( p = 0 . 006 ) , SOX2OT ( p = 0 . 0002 ) , and PSARL ( p = 0 . 0005 ) gene regions . MFN1 and PSARL both interact with OPA1 to regulate mitochondrial fusion and the inhibition of mitochondrial-led apoptosis , respectively . That two mitochondrial regulatory processes in the retina are implicated in the aetiology of myopia is surprising and is likely to provide novel insight into the molecular genetic basis of common myopia .
Myopia is the most common eye disorder , affecting an estimated 36% of adults over 20 years in the United States [1] and up to 61% in East Asia [2] . Myopia is a significant cause of vision loss [3] , and is becoming the most common single cause of blindness in the working age population [4] . Refractive error , measured in spherical equivalent ( SE ) diopters , is a quantitative trait influenced by multiple genetic and environmental factors . Myopia develops as a result of structural changes in the eye , particularly ocular axial length elongation , causing parallel rays of light to be focused in front of the retina , forming a blurred image . There are animal models for myopia development [5] , but the mechanisms responsible for detecting lack of focus and the signalling pathways from the retina to the choroid and ultimately to the sclera to induce eye growth , are not well-understood . Epidemiological studies have identified close visual work ( and correlates of this such as hours spent reading , education and IQ ) to be a significant risk factor for myopia development in children and that outdoor activity appears to be protective [2] . Twin studies also consistently demonstrate a large heritability for individual variation in refractive error around a specified population mean , ranging from 75–94% [6] . We previously described a genome-wide linkage analysis using autorefractor data for 221 dizygotic ( DZ ) female twin pairs , which identified 4 possible susceptibility loci , MYP7 on chromosome 11p13 , MYP8 on chromosome 3q26 , MYP9 on chromosome 4q12 , and MYP10 on chromosome 8p23 [7] . However , to date , these loci have not been replicated , and no known myopia susceptibility genes have been identified [8] . The aims of this study were twofold . The first was to replicate the linkage signals at these four loci using an independent sample of DZ twins to the original study , using measures of refractive error ( optician prescription ) obtained via a postal questionnaire . The second aim was to conduct a follow-up association study of the genomic region with strongest evidence of replicated linkage , using linkage disequilibrium mapping to identify possible susceptibility genes and to replicate results using an independent sample .
Refraction data , either from autorefractor or postal prescription , were available for 4273 UK twin subjects ( 1716 complete pairs ) with SE data . Overall , the SE mean ( between sib-pair standard deviation ) was −0 . 29D ( 2 . 36 ) , range −20D to +8 . 75D with an inter-quartile range of −1 . 06D to +1 . 125D and 26% of the subjects were myopic using a threshold of SE< = −1D . A total 1846 autorefractor measures ( 915 complete pairs ) and 997 postal prescriptions ( 485 pairs ) were available for the discovery phase of this study and an independent sample of 1430 twins for replication ( Table 1 ) . The mean age of subjects was 53 . 7 years ( SD 13 . 1 ) , range 16–82 years , and 90 . 3% of subjects were female . The high proportion of women is the result of long-term recruitment of female volunteers for study of phenotypes such as osteoporosis . For this study , we attempted to replicate linkage to four loci previously reported by our group [7] . Figure 1 illustrates linkage peaks to 3q26 for a discovery sample using autorefractive data with LOD 3 . 7 ( DZ twin pairs = 221 ) based upon previously published data [7] and replication sample using postal prescription data with LOD 2 . 12 ( DZ pairs = 485 ) . Combined linkage using pooled data gave LOD 2 . 63 ( DZ pairs = 706 ) . Marginal evidence for replicated linkage using the original Généthon map and microsatellite data for independent samples was also observed for MYP7 ( 11p13 ) and MYP9 ( 4q26 ) , but not for MYP10 ( 8p23 ) . These loci are currently subject to further investigation . Autorefractor rather than postal prescription data were used for discovery stage association mapping , since autorefractor data are observed to be more precise with a smaller standard deviation ( AR total sib-pair SD = 2 . 48; postal SD = 2 . 76; p = 0 . 0006 ) . Subjects measured using an autorefractor were measured in the same standardized manner with no transcription errors that tend to be associated with postal prescription data . For the initial fine mapping study of the 3q26 region ( Map 1 ) , a total of 243 cases and 257 controls selected from myopic and hyperopic concordant sib-pairs , respectively , were defined from the lower and upper quartiles of the SE quantitative trait using 915 twin pairs with complete autorefractor data ( see Materials and Methods ) . Seventy-nine of these had depleted DNA ( 46 with none and 33 samples with poor DNA quality or case-wise missing > = 30% ) , leaving 205 cases ( myopic individuals with a myopic sibling ) and 216 controls ( hyperopic individuals with a hyperopic sibling ) , a total of 421 case-controls for the preliminary Map 1 analysis ( Table 1B ) . The Map 2 data contained 154 new samples and approximately 70% of the samples from Map 1 , yielding a total of 443 case-controls . Hence a total of 575 cases and controls were genotyped for either the Map 1 ( n = 421 ) or Map 2 ( n = 443 ) discovery samples ( Table 1B ) , with 289 samples genotyped for both . It was intended to genotype the same samples for Maps 1 and 2 , but due to low DNA stock for some of the original Map 1 samples sent to Ellipsis for genotyping; new case/control samples with sufficient DNA stock were used to replace depleted Map 1 samples . Figure 2 illustrates the high-resolution linkage disequilibrium unit ( LDU ) map ( Figure 2A ) , based upon 24 , 331 HapMap PHASE II SNPs for the 3q26 region ( described in Materials and Methods ) , used to select informative markers for this study . The figure plots the relationship between cumulative genetic distance on the Y-axis ( LDU ) and physical location on the X-axis ( kb ) . The LDU map provides detailed information on fine-scale linkage disequilibrium . The horizontal steps seen in Figures 2B and 2C represent regions of extended LD , while rapid increments in cumulative LDU represent regions of breakdown in LD , primarily due to recombination [9] . The LDU map for the entire 3q26 region used for this study is presented in Table S1 . For the Map 1 samples , we attempted to genotype a total of 2304 SNPs . After removing non-polymorphic SNPs ( 384 SNPs ) , SNPs with a call rate ≤90% ( 84 ) , evidence of Hardy-Weinberg disequilibrium ( 38 ) and MAF< = 1% ( 0 ) , a total of 1800 out of 1920 polymorphic SNPs remained for Map 1 analysis . For the second stage of the association study ( Map 2 ) , in order to further refine the location of detected association , we genotyped an additional set of 382 SNPs for those LDU regions from Map 1 that showed evidence of association with myopic case-control status . Hence Map 2 had a high local LD resolution . After removing non-polymorphic Map 2 SNPs ( 33 ) , SNPs with a call rate ≤90% ( 19 ) , evidence of Hardy-Weinberg disequilibrium ( 20 ) and MAF< = 1% ( 7 ) , a total of 307 SNPs remained for Map 2 analysis . Map 1 provided preliminary multi-marker evidence of association to six gene regions ( Figure 2 ) , namely MYNN ( p = 0 . 0028 ) , MFN1 ( p = 0 . 02 ) , upstream of SOX2OT ( p = 0 . 0082 ) , a “gene desert” region downstream from SOX2OT ( between SOX2 and AT11B; p = 0 . 009 ) , MCF2L2/PSARL ( p = 0 . 014 ) and LPP ( p = 0 . 003 ) , with each analytical LDU window spanning 939 kb , 672 kb , 965 kb , 753 kb , 944 kb and 465 kb , respectively . Additional Map 2 genotyping within Map 1 provided the same or increased evidence of association for the MYNN ( p = 9 . 2×10−5 ) , MFN1 ( p = 1 . 54×10−8 ) , upstream of SOX2OT ( p = 1 . 1×10−11 ) , downstream of SOX2OT ( p = 1 . 6×10−5 ) and MCF2L2/PSARL regions ( p = 0 . 01 ) , but not LPP ( p = 0 . 03 ) . Hence four of the regions provided statistically significant evidence of association at the discovery phase with a significance threshold of α = 10−4 ( accounting for discovery multiple testing , see Materials and Methods ) , with the MFN1 and upstream SOX2OT regions attaining genome-wide significance ( α≈10−8 ) . For replication , we used an opportunistic sample in which we excluded all discovery twin samples ( and their co-twins ) from the TwinsUK register , to obtain 1430 individuals complete for autorefractor or postal SE and genotypes at 3q26 based upon the Illumina genome-wide Hap300 chip made available from other ongoing studies . Using quantitative tests of association , the same Malecot models and analytical LDU windows were fitted to the replication data . A significance threshold of α = 10−2 was used for the replication tests ( see Materials and Methods ) . All single-SNP allelic tests of association results ( see Association Mapping , Materials and Methods ) for Map 1 , Map 2 and replication samples are presented in Tables S2 , S3 , and S4 , respectively . Based on the Malecot model , the most likely physical location for a putative common functional variant in the MFN1 region was estimated to be at 180 , 566 kb with the 95% confidence interval ranging from 180 , 505–180 , 655 kb ( Map 2 , Table 2 ) . The variant location estimate at 180 , 566 kb lies in exon 7 of the MFN1 gene ( 180 , 548–180 , 594 kb , approximately 45 . 5 kb in length ) , but the confidence interval for this estimated location also includes the genes ZNF639 ( ZASC1 ) , MFN1 and GNB4 ( Figure 3 ) . Individual SNPs that showed strongest evidence of association for this window were rs6794192 ( 180 , 510 , 506 bp ) , rs10460887 ( 180 , 538 , 836 bp ) , rs9822116 ( 180 , 557 , 316 bp ) , rs17293193 ( 180 , 606 , 558 bp ) and rs7618348 ( 180 , 627 , 432 bp; all p-values provided in Table S1 ) . All five SNPs gave low p-values ( p<10−3 ) for single-SNP tests of association , with SNPs rs6794192 and rs7618348 genotyped and providing low p-values for all three samples ( Map 1 , Map 2 and replication ) and combined sample single-SNP p-values of 10−3 and 10−4 , respectively ( Table 2 ) . An annotated pair-wise LD plot is also presented for this region in Figure S1 . The MFN1 gene region result replicated for the independent sample of 1430 twins ( χ21 = 7 . 6 , p = 0 . 006 ) using a quantitative test of association , the same analytic LDU window and a different panel of markers for the 3q26 region derived from the Hap300 chip . Analysis of the Map 1 data yielded a significant window that covered the SOX2OT gene region ( χ21 = 7 . 0 , Table 3 ) . Further analysis using the Map 2 data showed a large increase in the significance level ( composite likelihood χ21 = 46 . 1 , p = 1 . 1×10−11 ) . The physical location for the putative associated common variant in the region using the more informative Map 2 data was estimated to be at 182 , 595 kb with a 95% confidence interval of 182 , 533–182 , 688 kb ( Table 3 ) . This location is approximately 156 kb upstream ( 5′ ) from the alternate-splicing ncRNA gene SOX2OT and 317 kb from the SOX2 transcription start sites ( Figure 4 ) . The confidence interval includes no known genes , but does include two predicted non-coding genes of unknown function ( floylorbu 0 . 51 kb in length and flerlorbu , 21 . 4 kb [10] and five putative alternative promoters upstream of SOX2OT , which between them cover a region of approximately 490 kb [11] . Individual SNPs most strongly associated with myopia for this window were rs1518933 ( 182 , 538 , 071 bp ) , rs733422 ( 182 , 604 , 752 bp ) and rs4855026 ( 182 , 609 , 663 bp ) . Figure S2 provides a pair-wise marker LD plot for the SOX2 gene region . These results were also observed for the replication sample using a quantitative test of association with Hap300 SNPs covering the same region ( χ21 = 14 , p = 1 . 8×10−4 , Table 3 ) . The SNP coverage for Map 2 did not include SNPs within or in close proximity of SOX2 ( Figure 4 ) , although the Map 2 evidence for association was based on an LDU analytical window that included the gene . The LDU maps illustrated in Figures 2C and 4 show evidence of multiple recombination hot spots around SOX2 . Preliminary marginal statistical evidence of association for Map 1 data was observed for the analytical LDU window at 184 , 313–185 , 257 kb ( χ21 = 6 . 0 , p = 0 . 014; Table 4 ) . This region has high recombination rates , is gene rich and includes the genes LAMP3 , MCF2L2 , B3GNT5 , KLHL6 , KLH24 , YEATS2 , MAP6D1 , PSARL , ABCC5 and HTR3D ( Figure 5 ) . Based on Map 2 data , the physical location for a putative causal variant was estimated to be at 184 , 386 kb with 95% confidence intervals 184 , 356–184 , 441 kb ( χ21 = 6 . 2 , p = 0 . 01; Table 4 ) . This location estimate lies within intron 3 of the 30-exon gene MCF2L2 with the confidence intervals including LAMP3 and MCF2L2 . An annotated pair-wise LD plot is presented for this region in Figure S3 . Strong evidence of association to this LDU window was also observed for the replication sample using a quantitative test of association for Illumina Hap300 SNPs genotyped for the same LDU window ( χ21 = 12 , p = 5×10−4 , Table 4 ) . However , the estimated location for a putative common causal variant for the same window and using Hap300 SNPs was different to that from the discovery SNP coverage ( Maps 1 and 2 ) . For Hap300 SNPs , the variant was estimated to be at 185 , 100 kb ( 95% CI 185 , 036–185 , 115 kb ) , located in the 3′ UTR of PSARL , with the confidence intervals including exons 4–10 of PSARL and the 5′ UTR of the neighbouring gene , ABCC5 ( Figure 5 ) . The statistical evidence for the PSARL location ( χ21 = 12 ) was stronger than that for MCF2L2 ( χ21 = 6 . 2 ) . The estimated locations for common functional variants at these three loci are presented in Table 5 . Evidence of association to these loci did not replicate using Hap300 samples , suggesting either Type 1 errors or failure to replicate due to the different genetic coverage of these regions provided by the Map 1 discovery and Hap300 marker sets .
Having first replicated the initial linkage to 3q26 , the strategy we adopted for fine mapping the large 30 Mb genomic region was to pursue evidence of association in two stages . First , using a high-resolution genetic map , we selected an informative set of SNP markers across the entire region , but at relatively low density to ensure economic feasibility . The second was to follow up those regions that showed strongest evidence of association in the region , with a denser set of markers placed on the same genetic map , on the assumption there are detectable common genetic variants in the region responsible for generating the observed linkage signal . The approach succeeded , with evidence of replicated association to the MFN1 , SOX2OT and PSARL gene regions . It is worth noting that the association initially detected in the three loci regions using Map 1 were only of marginal significance ( at p = 0 . 02 , p = 0 . 008 and p = 0 . 014 , respectively ) . However , when a higher-resolution map was genotyped for the locus , association was detected at genome-wide significance for MFN1 and SOX2OT . Evidence of association to the MCF2L2/PSARL gene region using Map 2 data remained the same ( p = 0 . 01 ) , but the same LDU window was subsequently replicated more strongly using a different panel of Hap300 SNPs ( p = 0 . 0005 ) . The diverging location estimates in the MCF2L2/PSARL region using two different SNP marker sets suggests the possibility of more than one common functional variant and co-incidental association for this LDU window ( Figure 5 ) . The latter emphasises how important informative SNP coverage is for detecting common variants and the use of marker panels that provide similar coverage of local LD patterns . The use of multi-marker tests can efficiently use the LDU locations to provide localization estimates , while for sparse marker sets the use of single SNP tests is likely to result in reduced power to detect association depending upon local LD structure . The results presented here are all the more remarkable in that we were able to replicate the same regions using an unselected sample , for a different panel of SNPs ( Hap300 ) genotyped at different centres . Some of the regions we have investigated on 3q26 are complicated with high recombination rates or a high density of genes . We have used a model that assumes common susceptibility loci with little or no allelic heterogeneity . As such we recognise there are likely to be more variants and genes in this region that will be identified and replicated by further mapping studies . The Malecot model delimits the MFN1 gene region ( using the most informative marker set , Map 2 ) with a 95% confidence interval ranging from 180 , 505–180 , 655 kb . Although the strongest evidence of association peaks at 180 , 565 . 8 kb in the middle of the MFN1 gene at exon 7 , the confidence interval includes two neighbouring genes , ZNF639 and GNB4 . Mitofusin-1 ( Mfn1 , the protein derived from MFN1 ) is a mitochondrial outer membrane protein , widely expressed in human tissues but varying in mRNA expression levels between tissues [12] . Mfn appears to be a key player in mediating mitochondrial fusion and morphology in mammalian cells [12] . Its interest as a possible candidate gene involved in ocular function stems from its relationship with OPA1 , a dynamin-related protein of the inner membrane which is mutated in autosomal dominant optic atrophy [13] , [14] . OPA1 requires Mfn1 to regulate mitochondrial fusion [15] . OPA1 is expressed in embryonic retina at many levels , not just the ganglion cells leading to the optic nerve , and continues to be expressed in adult retina with unknown function [16] . GNB4 is of interest as a myopia susceptibility gene , as the Gβ4 protein has been shown to be expressed in retinal ON bipolar cells . The function of bipolar cells in the retina is detection of the edge of objects , and so these cells may be involved in detection of hyperopic blur that is believed to drive the signal for eye growth in myopia . Inhibition of the retinal ON bipolar cells stops the compensatory eye growth when a negative lens or occluder is placed over chick or kitten eyes [17] . SOX2 is a fundamental homeobox gene , 2 kb in length , involved in ocular development , with mutations leading to anophthalmos [18] . The known interaction between the SOX2 and PAX6 genes in lens development suggests the possibility that these may also influence development of refractive error . PAX6 lies at the centre of our 11p13 linkage signal from the original linkage scan , although we found no intra-genic association with PAX6 using tagging SNPs [7] . Recent studies illustrate the important role that gene regulatory elements can play in disease susceptibility including for example , a homeobox transcription factor that influences heart development and subsequent risk of atrial fibrillation [19] . There is considerable body of evidence for the role of regulatory elements associated with PAX6 [20] , and on regulatory regions for SOX2 [21] . SOX2 itself lies in the intron of another larger ( 240 kb ) non-coding RNA gene SOX2OT , which may play a regulatory role in SOX2 expression [21] . SOX2OT is a highly complex locus , which appears to produce several proteins with no sequence overlap , with 14 documented alternative splicing mRNAs , 5 non-over-lapping alternate last exons and 7 validated alternative polyadenylation sites . Upstream of SOX2OT there are also 5 possible alternative promoters [11] ( DA281835–DA310380 ) and two putative ncRNA genes of unknown function , flerlorbu and floylorbu ( Figure 4 ) . Whether these elements co-operate with SOX2OT in regulating SOX2 is unknown . The protein presenilin-associated rhomboid-like protein ( PARL , coded for by the gene PSARL ) is a mitochondrial inner membrane protease , which interacts with OPA1 to inhibit the mitochrondrial remodelling process that signals apoptosis [22] . This reflects the broader phenomenon that molecular mechanisms behind mitochondrial morphology have been recruited to govern novel functions , such as development , calcium signalling , and apoptosis [23] . PARL plays two important known roles . The enzyme cleaves OPA1 to produce the anti-apoptotic truncated soluble form of OPA1 , which prevents cristae remodelling and the subsequent release of mitochondrial cytochrome c into the cytosol to stimulate apoptosis . The anti-apoptotic effects of these proteins are independent of mitochondrial fusion [22] . In addition , PARL appears to be implicated in mitochondria-to-nucleus signal transduction - following proteolytic processing of PARL , a small peptide sub-unit ( P-beta domain ) is released and translocated to the nucleus by an unknown mechanism [24] . The OPA1 gene encodes a 960 amino acid mitochondrial dynamin-related guanosine triphosphatase ( GTPase ) protein , which is transported from the nucleus to the outer surface of the inner mitochondrial membrane and interacts with Mfn1 and PARL to cause mitochondrial fusion and suppress mitochondrial-led apoptosis , respectively . OPA1 is widely expressed throughout the body , but most abundantly in the retina , followed by the brain . In the eye , OPA1 is present in the cells of the retinal ganglion cell layer , inner and outer plexiform layers and inner nuclear layer [16] . In summary , we have detected and replicated three novel loci at MFN1 , SOX2OT and PSARL using a multi-marker approach that models LD structure . We performed a two-stage design to ensure adequate SNP coverage using a high-resolution LDU map . Prior evidence of replicated linkage to this region means these associations are likely to be real with MFN1 , GNB4 , PSARL genes and regulatory non-coding RNAs in the vicinity of SOX2 , all plausible candidates . Although the mechanisms are not clear , this study strongly suggests that two fundamental mitochondrial molecular pathways are implicated in the aetiology of myopia . We are confident that additional mapping studies for these data are likely to replicate further candidate genes at 3q26-28 and genome-wide , which along with MFN1 and PSARL , can be taken forward to clarify the molecular genetic aetiology of common myopia .
Twins in this study volunteered through media campaigns to be on the TwinsUK Adult Twin Registry at St Thomas' Hospital , London [25] . Subjects were invited to attend the hospital for a visit , which involved collection of multiple phenotypes including measurement of refractive error using non-cycloplegic autorefraction ( ARM-10 autorefractor , Takagi Seiko , Japan ) , as well as venepuncture for blood collection for DNA extraction . For all studies , full informed consent was obtained , and protocols were reviewed by the Local Research Ethics Committee . A postal enquiry was also initiated in 2002–2003 , asking about ocular history and requesting subjects' ocular refraction prescription from their optometrist . We used postal data for those subjects without autorefraction data . Included in the questionnaires were questions about spectacle wear to cross-check refractive error data supplied . Subjects were excluded if they gave a history of cataract surgery , laser refractive surgery , retinal detachment or other ocular problems that might have influenced refractive correction . Spherical equivalent ( SE ) was recorded in the standard manner as the sum of the spherical power and half the cylindrical power in diopters ( D ) . The mean SE for left and right eye was calculated for each individual , and where data was available for only one eye , this was used as the SE for the subject . Although there has been little evidence of population stratification in population-based studies of self-reported Britons , we assessed for possible stratification with little or no evidence of stratification observed for these data [26] . We attempted to replicate the evidence from our original genome-wide analysis using AR data [7] for linkage to chromosomes 11p13 ( MYP7 ) , 3q26 ( MYP8 ) , 4q12 ( MYP9 ) and 8p23 ( MYP10 ) using the same ABI prism microsatellite marker set and Généthon genetic map . Refraction data for an independent sample of 485 DZ twin pairs was obtained from the postal questionnaire refraction data described above . Multipoint genome-wide linkage analyses were performed by use of the unadjusted mean SE of both eyes ( in D ) and optimal Haseman-Elston regression methods , implemented by use of a generalized linear model [27] . The most informative individuals were selected for genotyping from the lower and upper quartiles of the continuous SE diopter distribution . We selected individuals from a dataset of 915 twin pairs with complete autorefractor data , of whom 431 were monozygotic and 484 DZ pairs , which included the 221 DZ autorefracted pairs from the original linkage study . From the 915 twin pairs , a total of 575 unrelated cases and controls were selected for the discovery sample – 255 monozygotic and 320 DZ singletons . To enrich for genetically informative cases and controls , individuals were selected if they were myopic and had a myopic twin ( a “super” case ) or alternatively , were hyperopic and had a hyperopic twin ( a “super” control ) . The most myopic individuals ( cases ) , with a diopter score of less than <−1 , were selected from each twin pair , where the pair mean was equal to or less than = <−0 . 75 diopters . Similarly , the more hyperopic individuals ( controls ) , with a diopter score of at least >+1 , were selected from twin pairs with a pair mean greater than >+1 diopters . This resulted in an ascertained sample , designed to differentially increase allele frequencies between cases and controls for disease susceptibility alleles that predispose individuals to develop myopia . We chose hyperopic rather than normal sighted controls as a strategy to increase power , on the assumption that the aetiology for myopia and hyperopia lie on a continuum between health and disease and that both share genetic risk mechanisms . The discovery data were analysed using case-control status , on the supposition that most of the information would be captured by affection status , but we also tested the case-control data for quantitative association using the original diopter measurements for the selected data . Case and control samples were simultaneously genotyped using the same platform and arbitrarily allocated to the same plates . Case-control status was independent of plate and well assignment ( data not shown ) . The LD maps [28] assign markers to locations in linkage disequilibrium units ( LDU ) that describe the underlying structure of LD in the form of a metric map with additive distances . A high-resolution LDU map for the whole of chromosome 3 was constructed using the CEU PHASE II data from the HapMap Project [29] . The resulting LDU map for 3q26 was used for this study , corresponding to the region with replicated evidence of linkage . The 659 LDU region corresponds to approximately 42 . 7 cM on the decode linkage map [30] , implying that for 3q26 , on average ∼15 LDU correspond to 1 cM . For the first part of the project ( Map 1 ) , we selected three to four SNPs per 1 LDU across the entire 30 Mb region . This yielded a SNP density of approximately 1 SNP per 15 kb . This selection scheme captured the block-step structure of the high-resolution LD map and ensured good coverage of the LD steps . Map 1 ( the entire 3q26 region ) was first partitioned into 51 non-overlapping windows based on the LDU map with a minimum length of 10 LDU per window and by default , not breaking LDU blocks . For the six out of 51 LDU windows showing strongest evidence of association for the Map 1 data ( Figure 2 ) , an additional 382 SNPs were genotyped to refine the location estimate ( Map 2 ) . Further to the Map 1 and Map 2 studies , we also examined an independent replication sample of 1430 individuals for quantitative association , from which discovery samples and their relatives were excluded . The replication sample was composed of 460 unrelated individual female monozygotic and 338 dizygotic twin singletons and 316 DZ twin pairs . Tests of association were calculated using robust standard errors ( clustered by family identifier ) to account for relatedness with samples complete for refraction error ( autorefractor or postal ) and 3q26 genotype data . SNP genotypes for the replication samples were derived from a genome-wide Illumina HumanHap 300 dataset made available at the Twin Research Unit from other studies [26] . To assess the validity of postal SE with AR measures , we compared 138 individuals with both types of measure . The correlation between AR and postal data was 0 . 93 , with a mean difference ( AR – postal ) of −0 . 241 ( standard deviation = 0 . 93 ) and no observed relationship between the differences and the means for the two measures ( p = 0 . 75 ) . Discovery Map 1 and Map 2 sample handling , DNA genotyping and genotype calls were performed by Ellipsis Biotherapeutics Corporation ( Toronto ) using an Illumina Beadstation . SNPs were screened for quality control before analysis and rejected if the marker showed strong evidence of Hardy Weinberg disequilibrium ( at a threshold of χ21> = 12 ) , SNP-wise missing rates greater than 10% or MAF≤0 . 01 . Samples with a total of more than 30% case-wise missing were also removed before analysis . For the 3q26 replication sample , all samples were typed using the Infinium assay ( Illumina , San Diego , USA ) with fully compatible SNP arrays , the Hap300 Duo , Hap300 , and Hap550 . Quality control measures taken for these data are detailed in [26] . Allelic tests of association were initially performed for each marker . The association measure , z , from the 2×2 table between the myopia phenotype ( 0 , 1 ) and the two alleles of each SNP marker were obtained for Map 1 and Map 2 as z = |D|/f ( 1−R ) , where D is the covariance between myopia-status and the marker alleles , f is the frequency of myopic individuals in the sample and R is the minor allele frequency [31] . The significance of each window ( or LDU region ) was tested using a composite likelihood approach that simultaneously combines information from all markers within each window [32] on the basis of the Malecot Model . For the ith SNP , the observed association zi has an expectation E ( zi ) estimated by the model as: E ( zi ) = ( 1−L ) Me−εΔ ( Si−S ) +L . The parameter M ( intercept ) reflects a monophyletic or polyphyletic origin of susceptibility alleles ( i . e . proportion of disease alleles transmitted from founders ) . The parameter L ( asymptote ) is the spurious association at long distance . The object of LD mapping is to estimate S , which is the estimated location of the putative disease gene in the map . The parameter ε measures the rate of exponential decline in association with distance and hence Si is the LDU location of the ith marker . The Kronecker Δ is used for map direction and assures a correct sign , where Δ = 1 if Si≥S or Δ = −1 if Si<S . Given the observed associations for zi , the Malecot parameters are estimated iteratively by combining information over all loci within a window . The composite likelihood is calculated as Λ = ∑ Ki [zi−E ( zi ) ]2 , where z and E ( z ) are the observed and expected association values , respectively , at the ith marker SNP . Their squared difference is weighted by an information index Ki , which is estimated as: Kz = χ21/z2 , where χ21 is the Pearson's χ21 from the 2×2 table ( myopia status by SNP alleles ) . Following Maniatis et al . [32] , we used two different sub-hypotheses of the model to test for evidence of association . The null hypothesis is model Null where M = 0 . The alternative model Full allows the estimation of both M and S . Hence the contrast between these two models tests for association to a region and for a disease determinant at location S . The difference in marker-density between Map 1 , Map 2 and the replication samples genotyped for Hap300 SNPs , was taken into account by the use of an F statistic with df1 and df2 degrees of freedom . The degrees of freedom df1 was the number of SNPs minus df2 parameters estimated in the Full model . The F-value was estimated as F ( df1 , df2 ) = [ ( ΛNull−ΛFull ) /df2]/ΛF/df1 . Subsequently , to facilitate model fit comparison between tests with different degrees of freedom , p-values from the F-statistic were converted to a χ21 ( full details of methods are presented in [32] ) . The 95% confidence interval ( CI ) for the estimated location Ŝ was obtained as: Ŝ±t SE , where t is the tabulated value of Student's t-test for df2 degrees of freedom and SE is the standard error of parameter Ŝ . Estimates of Ŝ in LDU were converted to kb by linear interpolation of the two flanking SNPs . Same procedures were used for the replication samples ( Illumina HumanHap 300 , 1430 individuals ) . However , as the Hap300 chip had been genotyped for a large number of unselected twins , for this analysis we used the quantitative phenotype instead of the case-control status . Therefore the composite likelihood was calculated using the observed regression coefficient ( bi ) for each SNP marker ( i ) and the expected E ( bi ) , which was estimated using the Malecot model for every ith distance in LDU . For the association mapping study we present nominal p-values that do not correct for multiple testing . We used the following thresholds to indicate statistical significance at each stage: For the original discovery sample [7] we used a threshold of LOD 3 . 2 ( α≈10−4 ) to indicate genome-wide significance . For evidence of linkage replication presented in this study , we lowered the threshold to LOD 2 ( α≈2×10−3 ) , since replicating a true initial linkage result for complex traits is recognized to be difficult due to upward bias in discovery sample estimates [33] . Based upon Map 1 results , we took forward six LDU windows for further genotyping ( Map 2 ) that corresponded to the six most statistically significant results . For Map 2 we used a threshold of α = 10−4 , which is conservative , since a Bonferoni correction would provide a threshold of α≈10−3 ( 0 . 05/51 ) based upon approximately 51 independent tests ( i . e . 51 analytical windows were used to span the 3q26 30 Mb region ) . Replication association ( SE quantitative trait; Hap300 SNPs ) We attempted to replicate the six analytic LDU windows , with each test window independent of one another . Hence we considered replication using a threshold of α = 10−2 ( 0 . 05/6 ) based upon a Bonferoni correction . The URLs for data software presented herein are as follows: HapMap , http://www . hapmap . org/ ( for HapMap data )
|
Successful gene mapping strategies for common disease continue to require careful consideration of basic study design with the advent of genome-wide association studies . Here , we take advantage of prior information that the heritability of the quantitative trait myopia in the general population is high and shows evidence of replicated linkage to chromosome 3q26 . Based on this , we conducted a fine map linkage disequilibrium association study for the region , using a high-resolution genetic map derived from population-based HapMap Phase II data . For analysis , we used efficient multi-locus tests of association using single nucleotide polymorphism markers genotyped for our sample data and placed on the genetic map measured in linkage disequilibrium units . We followed up preliminary evidence of association for the discovery samples with further genotyping in the same samples to improve the model location estimates for the common functional variants we identified . Three locations were replicated using an independent sample . Two of the identified genes are likely to play an unexpected role in myopia with both pivotal in the healthy housekeeping metabolism of retinal mitochondria . Both proteins interact with OPA1 , with nonsynonymous OPA1 mutations causing the unrelated Mendelian disease Autosomal Dominant Optic Atrophy ( ADOA ) by triggering mitochondrial-led retinal ganglia cell apoptosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/gene",
"discovery",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"cell",
"biology/cell",
"adhesion",
"genetics",
"and",
"genomics/medical",
"genetics"
] |
2008
|
Identification and Replication of Three Novel Myopia Common Susceptibility Gene Loci on Chromosome 3q26 using Linkage and Linkage Disequilibrium Mapping
|
MicroRNAs ( miRNAs ) are ∼21 nt small RNAs that regulate gene expression in animals and plants . They can be grouped into families comprising different genes encoding similar or identical mature miRNAs . Several miRNA families are deeply conserved in plant lineages and regulate key aspects of plant development , hormone signaling , and stress response . The ancient miRNA miR396 regulates conserved targets belonging to the GROWTH-REGULATING FACTOR ( GRF ) family of transcription factors , which are known to control cell proliferation in Arabidopsis leaves . In this work , we characterized the regulation of an additional target for miR396 , the transcription factor bHLH74 , that is necessary for Arabidopsis normal development . bHLH74 homologs with a miR396 target site could only be detected in the sister families Brassicaceae and Cleomaceae . Still , bHLH74 repression by miR396 is required for margin and vein pattern formation of Arabidopsis leaves . MiR396 contributes to the spatio-temporal regulation of GRF and bHLH74 expression during leaf development . Furthermore , a survey of miR396 sequences in different species showed variations in the 5′ portion of the miRNA , a region known to be important for miRNA activity . Analysis of different miR396 variants in Arabidopsis thaliana revealed that they have an enhanced activity toward GRF transcription factors . The interaction between the GRF target site and miR396 has a bulge between positions 7 and 8 of the miRNA . Our data indicate that such bulge modulates the strength of the miR396-mediated repression and that this modulation is essential to shape the precise spatio-temporal pattern of GRF2 expression . The results show that ancient miRNAs can regulate conserved targets with varied efficiency in different species , and we further propose that they could acquire new targets whose control might also be biologically relevant .
MicroRNAs ( miRNAs ) are small RNA molecules , ∼21 nt in length , that are widespread regulators of gene expression in animals and plants [reviewed in [1] , [2]] . They recognize target RNAs by base complementarity and guide them to cleavage or translational arrest [1] . MiRNA-encoding genes are transcribed as primary transcripts harboring a fold-back structure with the miRNA embedded in one of its arms . In plants , these precursors are processed in the nucleus by the ribonuclease III DICER-LIKE1 ( DCL1 ) together with accessory components [reviewed in [3]] . The outcome of this processing activity is a miRNA/miRNA* duplex , which is 2′-O-methylated by HEN1 and incorporated into an AGO complex [reviewed in [1]] . The current version of the miRNA database ( miRBase 17 . 0 ) states the existence of more than 200 MIRNAs in Arabidopsis thaliana [4] . In many cases , MIRNA-genes can be grouped into families comprising different loci encoding similar or identical mature miRNAs [5] . However , the majority of the miRNAs in Arabidopsis are single young molecules indicating that their generation is a frequent process [6] , [7] , [8] , [9] , [10] . It is unclear whether these recently evolved miRNAs have a relevant biological contribution [11] , [12] . One exception might be the regulation of AGL16 by the young miR824 , which participates in stomatal development in Arabidopsis [13] . On the other hand , twenty-one miRNAs families are conserved in angiosperms , with some of them even present in lycopods and bryophytes [9] , [11] , [14] . In contrast to the younger ones , conserved miRNAs regulate key aspects of plant development , hormone signaling and stress response [11] . While plant miRNAs have extensive base-pairing to their targets , the interaction along the 5′ portion of the miRNA is the most relevant feature for its activity [15] , [16] . Interestingly , variations in the sequence of the small RNA among family members can have consequences on miRNA specificity [17] . In Arabidopsis , the miR159/miR319 family of miRNAs comprises six small RNAs that share 17 out of their 21 nt and regulate transcription factors of the TCP and MYB classes [16] , [17] , [18] , [19] , [20] . While miR319 can guide both types of targets to cleavage [17] , miR159 can only affect the MYBs [16] , [17] , [19] . MiRNA miR396 regulates GROWTH-REGULATING FACTORs ( GRFs ) [21] , [22] , [23] , [24] , a plant specific family of transcription factors known to be involved in the control of cell proliferation during leaf development [22] , [24] , [25] , [26] , [27] , [28] . The interaction between miR396 and the GRFs is unusual in plants as it contains a bulge in the 5′ region [22] , [23] . MiR396 accumulates with leaf age and restricts the pattern of expression of the GRFs to the proliferative region of the organ [22] . Overexpression of the miRNA causes a drastic reduction in cell number , while abolishing the regulation of GRF2 by miR396 promotes a moderate increase in organ size [22] . Interestingly , variations in the miR396 sequence can be found in different species ( miRBase 17 . 0 ) , such as a base insertion in the 5′ region of the miRNA , which is found only in rice and other monocotyledons [29] , [30] . It has been observed since the discovery of plant miRNAs that ancient miRNAs usually recognize similarly conserved target genes [16] , [23] , [31] , [32] . The occurrence of non-conserved targets recently incorporated during evolution to pre-existing miRNA networks has not been systematically studied so far , and more importantly , it is unknown whether the regulation of newly acquired targets has biological significance . Furthermore , miRNAs can have variations in their mature sequences in different species , which could potentially lead to neo-functionalization of the small RNA regulatory networks . Here , we characterize the miR396 regulatory network . We demonstrate its expansion to regulate a new target in species related to Arabidopsis thaliana and provide evidence pointing to the importance of this regulation in Arabidopsis development . We also show that monocot-specific variants of miR396 display an enhanced activity towards the conserved GRF transcription factors , while the sub-optimal regulation of the GRFs by miR396 in Arabidopsis might be important to quantitatively control GRF levels .
GRF regulation by miR396 is conserved at least in angiosperms and gymnosperms based on the presence of the small RNA [14] , [33] and GRF transcription factors harboring the miR396 target site ( Figure 1A ) . We began analyzing the existence of additional miR396 targets by searching the rice , poplar and Arabidopsis genomes using empirically-derived miRNA-target rules [16] . Transcription factors of the GRF class are the only conserved targets among these species ( see Tables S1 , S2 , S3 ) [23] , and their regulation by miR396 is known to be relevant for Arabidopsis development [22] . We also observed that 17 , 26 and 12 additional potential target genes that do not encode GRF transcription factors are predicted in Arabidopsis , poplar and rice , respectively ( see Tables S1 , S2 , S3 ) . Detection of RNAs cleaved at positions 10–11 with respect to cognate small RNAs is a hallmark of miRNA activity [34] , however , these products do not necessarily indicate a biologically relevant process . Therefore , we studied additional and potential miR396 targets from several points of view to identify those targets whose regulation would be more likely to have biological importance ( Figure 1B ) . Similar integrated strategies have previously allowed the identification of ta-siRNA targets [32] . First , we analyzed the expression of the 17 predicted targets in mutants of the miRNA biogenesis pathway in Arabidopsis using published ATH1 microarray data [32] , [35] . We found that two genes , At5g24660 and At1g10120 were up-regulated at least 30% in the miRNA biogenesis mutants hyl1 , serrate ( se ) and dcl1 ( Figure 1C; see Table S1 ) . This increase was similar to that observed for the miR396-regulated GRFs ( see Table S1 ) . At1g10120 analysis by a modified 5′ RACE-PCR revealed mRNA fragments consistent with a miR396-guided cleavage ( Figure 1D ) in agreement with previous results obtained for this gene by genome-wide analysis of miRNA targets [36] , [37] . In contrast , we did not find any pattern of miR396 activity on At5g24660 ( Figure 1E ) . We then performed a RT-qPCR with oligos spanning miR396 cognate sites in several siRNA and miRNA biogenesis mutants ( Figure 1F ) . At1g10120 was up-regulated two-fold in ago1 , se and hyl1 mutants , while no change was detected in dcl2-4 or rdr6 mutants ( Figure 1F ) . At5g24660 showed a moderate increase only in se and hyl1 mutants ( Figure 1F ) , albeit to a lower extent than that observed for At1g10120 . Finally , we prepared transgenic plants expressing an artificial target mimic directed against miR396 ( MIM396 ) to decrease the endogenous miRNA activity . These plants did not have any obvious phenotypic defects , similar to a previously described MIM396 prepared along a collection of target-mimics [12] , probably due to remaining miR396 activity . However , at the molecular level , we found that At1g10120 was up-regulated two-fold in 35S:MIM396 plants , which was analogous to the increase observed in miR396-regulated GRFs ( Figure 1G ) . In contrast , transcript levels of At5g24660 and GRF5 , which is not regulated by miR396 , were unaffected by the expression of MIM396 ( Figure 1G ) . Taken together , these results indicate that miR396 has a meaningful impact on the RNA regulation of this new target and that this regulation is quantitatively similar to the one observed in conserved GRFs . At1g10120 encodes a basic Helix-Loop-Helix ( bHLH ) transcription factor , namely bHLH74 [38] , [39] , [40] . Interestingly , we observed that the miR396-binding site of bHLH74 is formed after the splicing of the first two exons ( Figure 1H ) . We searched for bHLH74 homologs with a miR396 target site in EST and genome sequence databases of species related to Arabidopsis thaliana . Our observations indicated that the miR396-bHLH74 regulatory module is present in Brassicaceae species . We also found that conservation extends to Cleome spinosa which belongs to the sister family Cleomaceae , separated 40–50 million years from Arabidopsis thaliana [41] , [42] ( Figure 1H; see Figure S1 and Table S4 ) . We did not find evidence of bHLH74 homologs with a miR396 target site in more distant species of Arabidopsis thaliana , either looking at syntenic regions of sequenced genomes such as poplar or by BLAST search against EST databases . The search was also performed trying to identify relaxed target sites , for example , looking only at the 5′ region which is completely located in the second exon of bHLH74 without intron interruption . In neither case did we find an additional bHLH74 homologue that could be potentially regulated by miR396 . The up-regulation of bHLH74 in miR396-deficient plants and its conservation in a group of related species suggested that miR396 regulation might have a biological significance . To study the importance of bHLH74 regulation by miR396 in more detail , we prepared a version of the gene with mutations that impaired its interaction with the miRNA ( rbHLH74 ) . We introduced silent mutations to avoid changing the encoded amino acids and did not modify the region next to the intron-exon junction ( Figure 2A ) . First , we prepared transgenic plants expressing the wild-type and miR396 resistant gene from the viral CaMV 35S promoter . We observed that both transgenes were able to cause developmental defects , however the effects caused by the overexpression of the miRNA resistant gene ( 35S:rbHLH74 ) were stronger and in the most extreme cases led to the formation of chlorotic seedlings , which failed to develop shoot apical meristems ( see Figure S2 ) . These results demonstrate that high levels of bHLH74 can be toxic for normal plant development and that miR396 can down-regulate bHLH74 expression levels in vivo . Next , two vectors expressing a genomic version of the transcription factor with different sensitivities to miR396 , including the endogenous upstream regulatory regions , were constructed . For an initial characterization of the resulting transgenic plants , we analyzed bHLH74 transcript levels in mature leaves . As miR396 accumulates with leaf age [22] , we expected large differences in bHLH74 mRNA abundance in these samples . The genomic version of the transcription factor containing silent mutations that impaired its regulation by miR396 ( bHLH74:rbHLH74 ) accumulated varied levels of mRNA reaching levels eighty-fold higher than those for the endogenous transcript in the most extreme cases ( Figure 2B ) . On the other hand , the wild-type version ( bHLH74:wtbHLH74 ) accumulated at most eight-fold more ( Figure 2B ) . These results are consistent with high levels of endogenous miR396 guiding the cleavage of the wild-type bHLH74 transcript . Note that the differences in mRNA accumulation between bHLH74:rbHLH74 and bHLH74:bHLH74 are smaller in younger developing tissues , where miR396 levels are low ( see below ) . These transgenic plants also displayed alterations in leaf development , especially in the vein pattern and organ shape , which had sharper edges than those of wild-type leaves ( Figure 2D–2E ) . The angle formed at the distal part of the leaf was significantly reduced in most bHLH74:rbHLH74 transgenics ( Figure 2C–2D ) . Furthermore , there was also a reduction in the number of branching points ( NBP ) in the vasculature [43] in plants with high bHLH74 levels ( Figure 2D , inset ) . These results show that bHLH74 regulation by miR396 might be biologically important for Arabidopsis development . The analysis of transgenic plants harboring a rbHLH74 transgene suggested that bHLH74 might play a role during Arabidopsis leaf development . To further explore this possibility , we identified a loss-of-function mutant for transcription factor bhlh74-1 ( Figure 3A ) . Determination of bHLH74 mRNA levels indicated that they were severely reduced in this mutant ( Figure 3B ) . As rbHLH74 caused a reduction in the number of branching points in the leaf vasculature , we analyzed NBP values for bhlh74 mutants . Interestingly , we found that the NBP was increased approximately 24% in bhlh74-1 with respect to the wild type ( Figure 3C–3D ) . This phenotype was opposite to that found in plants harboring a rbHLH74 transgene . Actually , we observed a quantitative response between the increase of rbHLH74 mRNA levels and the reduction of NBP ( Figure 2B and Figure 3C–3D ) . Together with previous reports [22] , these results show that both types of miR396 targets , namely GRFs and bHLH74 , have biological roles during leaf development . We then compared the regulation of the new bHLH74 target by miR396 to the regulation of an ancient GRF target , such as GRF2 . First , we prepared a reporter to follow miR396 expression by fusing the 2 Kb upstream regulatory sequences of MIR396b to a GUS reporter . The miRNA reporter increased its expression during leaf development ( Figure 4A ) . In young leaves MIR396b:GUS was expressed in a gradient along the longitudinal axis of the leaf , with higher expression at the distal part ( Figure 4A ) . At later developing stages , MIR396b:GUS was detected in whole organs ( Figure 4A ) . The profile of the reporter matched previous small RNA blots performed for this miRNA [22] . A wild-type reporter for GRF2 containing the upstream regulatory region as well as the first four exons harboring the miR396 target site was expressed in young leaves and in proximo-distal gradient along the longitudinal axis of the organ ( Figure 4B ) [22] . Interestingly , the pattern of MIR396b and GRF2 expression was complementary during leaf development . This was especially noticeable in 15-day old seedlings where MIR396b was expressed in older organs and in the distal part of young leaves , while GRF2 was expressed in the proximal part of young organs ( Figure 4A–4B ) . A miR396-resistant GRF2 reporter , with mutations in the target site , was expressed in a broader domain , highlighting the activity of the miRNA in shaping GRF2 expression ( Figure 4B ) [22] . We then prepared two reporters to follow the regulation of bHLH74 by miR396 . We fused its promoter and the first two exons , which generate the miRNA target site , to the GUS gene ( Figure 4C ) . This first sensor , which has a functional miR396-binding site ( wtbHLH74-GUS ) was strongly expressed in young leaves , especially in the veins , while its expression decreased in older leaves ( 10 out of 16 independent transgenic plants ) ( Figure 4C ) . The second sensor contained silent mutations in the miRNA binding site , which impaired its regulation by miR396 ( rbHLH74-GUS ) . This reporter was expressed in organs much older than those of the wild-type version highlighting the role of miR396 in restricting its activity to younger organs ( 12 out of 16 independent transgenic plants ) ( Figure 4C ) . Although we cannot disregard the possibility of the existence of other regulatory levels affecting bHLH74 expression such as post-translational modifications , which are not detected by our sensors , the results show that miR396 contributes to the spatio-temporal regulation of bHLH74 . Furthermore , the expression of the bHLH74 sensors in the veins is in agreement with the biological role of the transcription factor in the control of leaf vasculature development . We then analyzed miR396 , GRF2 and bHLH74 transcript levels by RT-qPCR in young and fully-expanded leaves ( Figure 4D ) . We observed that while miR396 was induced several times during leaf development , both GRF2 and bHLH74 decreased significantly . These quantitative measurements are in accordance with the whole-mount GUS stainings , supporting the function of miR396 in the regulation of both types of targets during organ growth ( Figure 4A–4D ) . Finally , we measured the accumulation of bHLH74 mRNA in transgenic plants expressing the wild-type and miR396 resistant version of bHLH74 at two leaf developmental stages . At young stages when miR396 levels are low , bHLH74 was only slightly higher in bHLH74:rbHLH74 than in bHLH74:wtbHLH74 transgenic plants ( Figure 4E ) . However , the wild-type bHLH74 was significantly down-regulated more than ten times in older organs , in agreement with the activation of miR396 expression ( Figure 4E ) . These results further support the role of miR396 in the regulation of bHLH74 expression . The fact that bHLH74 regulation by miR396 is important for Arabidopsis development suggests that the miR396 regulatory network could be relatively dynamic at least with respect to the acquisition of new targets . We then explored whether there could be biologically relevant variations in the miRNA . Analysis of miR396 variants in different species using the miRNA database ( miRBase 17 . 0 ) indicates that there are , indeed , several variants of miR396 ( Figure 5 ) . In Selaginella , miR396 has a G at position 7 , while both genes in Arabidopsis encode small RNAs with an A at that position ( Figure 5A ) . Pine and poplar have precursors for both types of mature miR396 species ( Figure 5A ) . Interestingly , there is a miR396 variant with an insertion of a G at position 7–8 of the miRNA ( Figure 5A ) . This variant was first detected in rice [29] , and further studies indicated that it is found only in monocotyledons ( Figure 5B ) [29] , [30] , [44] . Sequence differences among related miRNAs could be important in plants as it has been shown for miR319 and miR159 , which have very similar sequences but regulate different genes [17] . Interestingly , the changes observed for the miR396 sequence at position 7–8 are predicted to have an important effect on miRNA activity based on previous biochemical data and mutant analysis [15] , [16] , [17] . We also searched for potential variations in other ancient miRNAs annotated in the miRNA database ( miRBase 17 . 0 ) , and found that changes in the 5′ region of the miRNA might also exist in other cases ( see Figure S3 ) . It should be considered that some of the polymorphisms observed could arise from non-expressed paralogs that have been annotated due to their homology to other known miRNAs . To confirm the expression of the miR396 sequences , we analyzed publicly available deep-sequencing small RNA libraries from several species [29] , [33] , [45] , [46] , [47] , [48] , [49] , [50] . We found that most small RNAs were detected in vivo , confirming that a complex spectrum of miR396 sequences co-exists in nature ( Figure 5B; see Table S5 ) . Interestingly , we observed that the monocot-specific variant was the most abundant miR396 variant in rice , maize and Brachypodium distachyon as judged by deep-sequencing ( Figure 5B ) . Recognition of the GRF target site by miR396 generates a bulge between positions 7 and 8 of the miRNA ( Figure 6A ) . The insertion of one nucleotide in the monocot-specific version of miR396 eliminates this bulge , strengthening the interaction of the miRNA-target pair by 7 kcal/mol ( Figure 6A–6B ) . Since the contribution of a bulge to the activity of a miRNA has not yet been assayed in plants , we decided to explore it in more detail . For this purpose , we expressed the two versions of miR396 from the MIR319a precursor in Arabidopsis , which has already been shown as an efficient driver of artificial miRNA sequences [51] . Ectopic expression of the Arabidopsis miR396 mature sequence from the MIR319a precursor caused smaller and lanceolated leaves ( Figure 6C–6F ) , in a similar way to the overexpression of the endogenous MIR396 precursor [21] , [22] . Expression of the monocot-version of miR396 ( miR396_7-8insG ) caused stronger effects on the leaf lamina ( Figure 6C–6F ) . We further expressed both miR396 variants from the endogenous MIR396b promoter . While an additional copy of the endogenous miR396 sequence caused a minor impact on leaf development , expression of the monocot miR396 version from the MIR396b promoter affected leaf development in more than 50% of the independent transgenic plants ( Figure 6C–6F ) . These results show that the miR396 monocot variant is more active in vivo . We then set up an assay to quantify both miR396 variants by RT-qPCR in the same reaction ( for details see Figure S4 ) . We focused on transgenic plants with a moderate reduction in leaf lamina ( ∼60% ) , which was observed in 52% or 38% of the primary transgenic plants expressing the endogenous or the monocot-specific miR396 sequence from the 35S viral promoter , respectively ( Figure 6D–6F ) . We measured the levels of miR396 in the transgenics over-accumulating the endogenous Arabidopsis miRNA and found that it was eight-fold higher compared to wild-type levels ( Figure 6G , for a small RNA blot see Figure S5 ) . In contrast , transgenic seedlings with the same phenotype but expressing the monocot-specific miR396 variant displayed only a two-fold increase ( Figure 6G ) . Analysis of the miR396-regulated gene GRF2 in seedlings with moderated phenotypes revealed a decrease of its mRNA levels to 40% in both types of transgenic plants ( Figure 6H ) . We also tested the activity of the two miRNA variants on bHLH74 . While the overexpression of the endogenous miR396 significantly down-regulated bHLH74 mRNA ( approximately 90% ) , the monocot-specific version had only a minor effect ( Figure 6I ) . Furthermore , bHLH74 transcript levels were reduced only 25% in plants expressing the highest levels of the monocot-specific version ( data not shown ) . Therefore , this monocot-specific version of miR396 is selectively more efficient towards the GRFs . The addition of an extra nucleotide to this variant causes a bulge to be formed on the miRNA side of the bHLH74/miR396 pair ( see Figure S6 ) . Altogether , these results show that asymmetric bulges located either in the miRNA or in the target dampen the miRNA-guided cleavage reaction . We also analyzed the activity of the miR396 variant found in Selaginella , pine and poplar ( Figure 4A ) and determined that it caused slightly stronger developmental defects than the wild-type precursor ( see Figure S7 ) . A possible explanation to this is that the miR396_7A>G version replaces an interaction between the A-U pair with a stronger G-C pair , causing a concomitant change of more than two kcal/mol in the interaction energy of the miR396/GRF pair ( see Figure S7 ) . The previous results show that a single miR396 gradient generates opposing gradients of expression for its targets ( Figure 4 ) , and that a perfect match between miR396 and the GRFs increase the in vivo efficiency of the miRNA ( Figure 6 ) . Next , we decided to analyze whether the bulge present in the miR396-GRF pair plays a role in patterning the expression of GRF2 during leaf development in Arabidopsis . To test this , we designed another GRF2-GUS reporter where the bulge was eliminated from the interaction with the endogenous miR396 , thus generating a nearly perfect pairing ( pGRF2-GUS ) ( Figure 7A; see Table S6 ) . Whole-mount stainings of wtGRF2-GUS revealed its expression in young developing leaves of Arabidopsis in 17 out of 20 primary transgenics ( Figure 7A ) . The pGRF2-GUS construct had a more limited expression and was restricted towards the proximal part of the leaf . This typical expression pattern was observed in 15 out of 20 transgenic plants for pGRF2-GUS ( Figure 7A ) . The remaining reporter lines displayed even weaker levels of GUS staining . We then measured the GUS mRNA in two representative lines for each vector and observed that bulge removal from the GRF2 reporter caused approximately a two-fold reduction in GUS RNA , confirming its higher sensitivity to miR396 ( Figure 7B ) . These results further support our previous findings , i . e . the monocot-specific version of miR396 , which does not have a bulge in the miRNA-target pair , has a higher activity towards the GRFs than the one from Arabidopsis . It has been shown that GRF2 is expressed in the proximal part of the leaf , which contains proliferating cells [22] . We then compared the expression pattern of several reporters in developing leaves of Arabidopsis thaliana . We observed that MIR396b has an expression gradient opposite to that of wtGRF2-GUS and pGRF2-GUS expression ( Figure 7C ) . However , the shape of the latter two gradients is different , being pGRF2-GUS tapered off faster than wtGRF2-GUS . As expected , rGRF2-GUS was expressed throughout the leaf ( Figure 7C ) . We also stained a CYCLINB1;1 reporter to identify proliferating cells in similar organs . Comparison of the expression patterns for GRF2 reporters revealed that only the one harboring the wild-type target sequence was co-expressed with the proliferating cells ( Figure 7C ) . Altogether , these results suggest that particular miRNA-target interactions , which are not a perfect match and therefore likely operate at sub-optimal activity , might have biological implications .
Most of the conserved miRNAs regulate transcription factors involved in development and hormone signaling . These target genes generally contain similarly conserved miRNA binding-sites . Studies performed in Arabidopsis thaliana and other species have shown that interfering with the regulation of conserved targets by changes in the recognition sites , mutations in the miRNAs [reviewed in [1] , [2]] or expressing miRNA-target mimics [12] usually lead to important developmental defects . Genome-wide analyses of miRNA targets have revealed that ancient miRNAs can regulate targets that are not broadly conserved [36] , [37] . However , detection of miRNA-guided cleavage is not necessarily indicative of a relevant biological function . This has already been pointed out for young miRNAs , whose biological functions remain unclear , even though their activities can be detected in vivo [9] , [11] . The results presented here show that the regulation of bHLH74 by miR396 has a meaningful impact on its RNA levels , in a similar way to that observed for the widely distributed GRF transcription factors . Interestingly , both targets are involved in leaf development . While the GRFs control cell number [21] , [22] , [24] , [26] , [28] , we found that bHLH74 regulates the vein patterning of the leaf . Furthermore , GRF2 [22] and bHLH74 transgenes harboring silent mutations in their miRNA target sites affect leaf development , suggesting that the regulation of both type of targets by miR396 is important for Arabidopsis development . GRF regulation by miR396 can be traced back to at least the gymnosperms based on the existence of GRF transcription factors with a miRNA target site . In contrast , we could only find bHLH74 homologues with miR396-binding sites in species within the sister families Cleomaceae and Brassicaceae . However , it is tempting to speculate that the conservation of a miRNA-target sequence in related species might still be significant and could serve as a tool to identify additional miRNA targets whose regulation might be of biological relevance . Ancient miRNAs are usually found as small gene families , encoding small RNAs of similar or identical sequences . One of the advantages of having families with multiple miRNA members is to provide flexibility in the way miRNAs are themselves regulated [52] . Additionally , differences in the miRNA sequences could cause related miRNAs to regulate different sets of targets . This has been previously shown for the miRNAs miR319 and miR159 , which are similar in sequence but still regulate different genes [16] , [17] , [18] , [19] . While miR319 can guide TCP and MYB genes to cleavage , specific base differences prevent miR159 activity on the TCPs [17] . The results reported here suggest that similar sequences could have also acquired specialized functions to regulate the same set of targets with different efficiency . The conserved interaction between miR396 and the GRFs has a bulge at position 7–8 of the target site . Removal of this bulge by either the addition of one base to the miRNA , as seen in the monocot-specific miRNA variant , or by the removal of a base from the GRF target site , results in a higher miRNA activity . Interestingly , the addition of a base in the Arabidopsis miR396 selectively improves its efficiency towards the GRFs at the expense of losing activity towards bHLH74 , suggesting that bulges in miRNA/target pairs could be used for differential target regulation in miRNA networks . Our systematic analysis of evolutionary conserved plant miRNAs has shown that additional variations exist in other miRNA sequences , including changes in the 5′ sensitive region . It might be interesting to further explore the occurrence in nature of additional examples of miRNA specialization , and whether they cause changes in target specificity and/or efficiency . It has been proposed that Arabidopsis miR396 contributes to the fine-tuning of GRF expression [22] . Here , we have further shown that only a GRF2 reporter under suboptimal regulation by the endogenous miR396 can overlap the proliferative region of a developing leaf in Arabidopsis thaliana . It is plausible , however , that a gross down-regulation of the GRFs under specific conditions or in specific cells , such as the one caused by the monocot miR396 variant , could also be advantageous . MiR396 is expressed at low levels in the meristem and leaf primordia , and then it steadily accumulates as leaves develop [22] . When considering a single developing organ , miR396 accumulates in the more mature and distal part , with a miRNA gradient proceeding towards the base of the organ . Analysis of bHLH74 and GRF2 expression patterns revealed that they both shared a temporal component , which is imposed by the accumulation of miR396 during leaf development . Both genes are expressed in young organs as a consequence of miR396 activity . Still , bHLH74 and GRF2 reporters do not have identical expression patterns , as the bHLH is more restricted to the vasculature while the GRF is more widely distributed throughout the leaf mesophyll cells . The exact tissue of expression might be governed by cis regulatory sequences in the promoters of the different targets . Therefore , several layers of regulation can ultimately generate unique and coordinated expression patterns on target genes belonging to the same miRNA regulatory network . Another potential level of complexity in the regulation of different targets by miR396 might arise from the miRNA expression gradient in a developing leaf , which extends from the distal part of the organ towards its base . In principle , a single miRNA gradient can generate different expression gradients of its targets , depending at least partially on the exact nature of the miRNA-target pair .
Arabidopsis ecotype Col-0 was used for all experiments , with the exception of the GRF2-GUS and bHLH74-GUS reporters , which were analyzed in rdr6 background . Plants were grown in long ( 16 h light/8 h dark ) or short photoperiods ( 8 h light/16 h dark ) at 23°C . To analyze the vein pattern , leaves were fixed with FAA and cleared with a chloral hydrate solution . Pictures were then taken under dark field illumination in a dissecting microscope . The number of branching points ( NBP ) [43] was measured per leaf half in Figure 2D , and in the whole leaf in Figure 3D . See Table S6 for a list of binary plasmids used in this study . The miRNA target motif in GRF2 and bHLH74 was altered introducing synonymous mutations in a cloned wild-type genomic fragment using the QuikChange Site Directed Mutagenesis Kit ( Stratagene ) . Artificial miRNAs [51] were generated by PCR , and MIM396 was generated by gene synthesis ( Mr . Gene GmbH ) . RNA was extracted using TRIzol reagent ( Invitrogen ) and 1 , 0 µg of total RNA was treated with RQ1 RNase-free Dnase ( Promega ) . Next , first-strand cDNA synthesis was carried out using SuperScriptTM III Reverse Transcriptase ( Invitrogen ) with the appropriate primers . PCR reactions were performed in a Mastercycler ep realplex thermal cycler ( Eppendorf ) using SYBRGreen I ( Roche ) to monitor dsDNA synthesis . MiR396 and miR396_7-8insG levels were concurrently determined in each sample by stem-loop RT-qPCR [53] . A scheme of the strategy used for the simultaneous quantification of miR396 and miR396_7-8insG is provided in Figure S4 . Relative transcript level was determined for each sample , normalized using PROTEIN PHOSPHATASE 2A ( AT1G13320 ) cDNA levels [54] . MiR396 levels were also estimated by small RNA blots as described previously [22] . Primer sequences are given in Table S7 . To visualize reporter activity , transgenic plants were subjected to GUS staining , as described previously [55] . RNA adaptor ligation , reverse transcription and 5′RACE were performed according to the procedure described previously in order to determine RNA degradation products [56] . Small RNA sequences obtained from miRBase ( 17 . 0 ) [4] were used for analyses of miRNA sequence variations in conserved miRNA families in angiosperms [9] . A consensus sequence was identified for each family and deviations from the consensus at each position were quantified . The number of variations was normalized to the total number of miRNA family members , so that each family contributed equally . A list of relevant AGI locus identifiers is provided in Table S7 .
|
Plants and other multicellular organisms need precise spatio-temporal control of gene expression , and this regulatory capacity depends , in part , on small RNAs . MicroRNAs ( miRNAs ) are one class of ∼21 nt small RNAs that originate from endogenous fold-back precursors found in plants and animals . They recognize complementary target sites in target mRNAs and guide them to cleavage or translational arrest . Studies of conserved miRNA networks in Arabidopsis and other plants have revealed that they fulfill essential regulatory roles . Most of the ancient miRNAs regulate transcription factors involved in plant development and hormone signaling . Here , we characterize the miR396 regulatory network . While miR396 regulates GRF transcription factors , at least in angiosperms and gymnosperms , this miRNA additionally regulates another transcription factor of the bHLH class but only in Arabidopsis thaliana and closely related species . Most conspicuously , the regulation of both conserved and new targets is important for leaf development in Arabidopsis . We also show that miRNA variants can exist in certain species and that they can display an enhanced activity towards their targets . In summary , we propose that conserved miRNA regulatory networks might expand their functions by the recruitment of additional targets as well as by slight variations in the small RNA sequences .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"science",
"rna",
"interference",
"plant",
"biology",
"gene",
"expression",
"plant",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"molecular",
"biology"
] |
2012
|
Functional Specialization of the Plant miR396 Regulatory Network through Distinct MicroRNA–Target Interactions
|
The Togavirus ( Alphavirus ) Mayaro virus ( MAYV ) was initially described in 1954 from Mayaro County ( Trinidad ) and has been responsible for outbreaks in South America and the Caribbean . Imported MAYV cases are on the rise , leading to invasion concerns similar to Chikungunya and Zika viruses . Little is known about the range of mosquito species that are competent MAYV vectors . We tested vector competence of 2 MAYV genotypes in laboratory strains of six mosquito species ( Aedes aegypti , Anopheles freeborni , An . gambiae , An . quadrimaculatus , An . stephensi , Culex quinquefasciatus ) . Ae . aegypti and Cx . quinquefasciatus were poor MAYV vectors , and had either poor or null infection and transmission rates at the tested viral challenge titers . In contrast , all Anopheles species were able to transmit MAYV , and 3 of the 4 species transmitted both genotypes . The Anopheles species tested are divergent and native to widely separated geographic regions ( Africa , Asia , North America ) , suggesting that Anopheles may be important in the invasion and spread of MAYV across diverse regions of the world .
Mayaro virus ( MAYV ) is a member of the genus Alphavirus ( family Togaviridae ) which was first isolated from the blood of five febrile workers in Mayaro County , Trinidad , in 1954 [1] . MAYV has a single-stranded positive-sense RNA genome of approximately 11 . 7 kb and is classified into three genotypes: D , L , and N [2 , 3] . Genotype D ( dispersed ) includes strains isolated in several South American countries , whereas genotype L ( limited ) includes strains isolated only in Brazil . In 2010 , a minor genotype called N ( new ) , was isolated in Peru , but it is limited to one known sequence . Since its first isolation , MAYV has caused sporadic outbreaks and small epidemics in several countries in South and Central America ( reviewed in [4] ) . In 2015 , an 8-year-old child from Haiti was reported as co-infected with MAYV and dengue virus ( DENV ) , suggested that MAYV may be actively circulating in the Caribbean [5] . Several imported cases recently reported in the Netherlands [6] , Germany [7] , France [8] , and Switzerland [9] highlight the need to survey naive regions , including the United States , for possible introductions of this neglected arthropod-borne virus ( arbovirus ) . The symptoms of MAYV infection include rash , fever , myalgia , retro-orbital pain , headache , diarrhea , and arthralgia , which may persist for months or even years [10] , and are similar to those caused by others arboviruses such DENV or chikungunya virus ( CHIKV ) . Due to the absence of routine differential diagnostics , reported cases of MAYV likely underestimate the real prevalence , and the circulation of the virus can pass undetected in areas with ongoing DENV or CHIKV outbreaks [4 , 11] . MAYV is thought to be principally transmitted by the bite of diurnal canopy-dwelling mosquitoes of the genus Haemagogus [4] . These mosquitoes are responsible for maintaining the sylvatic cycle involving nonhuman primates and birds as primary and secondary hosts , respectively . Human infections are sporadic , likely because Haemagogus spp . rarely display anthropophilic behaviors , and they possess a preference for rural areas with proximity to forests [12] . Vector competence ( VC ) studies demonstrated that anthropophilic and urban-adapted species , such as Aedes aegypti ( Linnaeus , 1762 ) and Ae . albopictus ( Skuse , 1895 ) , are competent vectors for MAYV in laboratory conditions [13–15] . Culex quinquefasciatus mosquitoes positive for MAYV have also been identified from field collections during a DENV outbreak in Mato Groso County , Brazil [16]; however , their capacity to transmit MAYV has not been demonstrated . Overall , little data is available about the VC of mosquitoes for MAYV [14 , 15 , 17 , 18] and , there have been no studies about the VC of vector species in the United States ( with the exception of two field populations of Ae . aegypti and Ae . albopictus from Florida [15] ) . To address this knowledge gap , we evaluated the ability for Ae . aegypti , Anopheles freeborni ( Aitken , 1939 ) , An . gambiae ( Giles , 1902 ) , An . quadrimaculatus ( Say , 1824 ) , An . stephensi ( Liston , 1901 ) and Cx . quinquefasciatus ( Say , 1823 ) to become infected with and transmit MAYV after feeding on a viremic blood meal . Our results demonstrate that while the laboratory strains of Ae . aegypti and Cx . quinquefasciatus tested are poor vectors for MAYV , all tested Anopheles species were competent laboratory vectors for MAYV , including species that they have the potential to support the transmission cycle if the virus is introduced into the United States . Additionally , the results of our study provide useful information to improve entomologic surveillance programs and prevent future outbreaks of this emerging neglected pathogen .
Six mosquito species were used in this experimental study . The An . gambiae ( NIH strain ) were originally obtained from The National Institutes of Health ( Bethesda , MD , USA ) . An . stephensi ( Liston strain ) were provided by Johns Hopkins University ( Baltimore , MD , USA ) . Cx . quinquefasciatus ( Benzon strain ) were provided by the Wadsworth Center ( Slingerlands , NY , USA ) and was initially derived from a colony maintained by Benzon Research ( Carlisle , PA , USA ) . An . quadrimaculatus ( Orlando strain , MRA-139 ) and An . freeborni ( F1 strain , MRA-130 ) were provided by BEI Resources ( Manassas , VA , USA ) . Ae . aegypti ( Rockefeller strain ) were provided by Johns Hopkins University . Anopheles species were selected to cover different geographical macroregions: North America ( An . freeborni and An . quadrimaculatus ) , Africa ( An . gambiae ) and Southeast Asia ( An . stephensi ) . Mosquito colonies were reared and maintained at the Millennium Sciences Complex insectary ( The Pennsylvania State University , University Park , PA , USA ) at 27°C ± 1°C , 12:12 h light:dark diurnal cycle at 80% relative humidity in 30×30×30-cm cages . Ground fish flakes ( TetraMin , Melle , Germany ) were used to feed Anopheles spp . and Aedes sp . larvae . A 1:1:1 mixture of bovine liver powder ( MP Biomedicals , Solon , OH , USA ) , koi pellets ( TetraPond Koi Vibrance; TetraPond , Prestons , Australia ) , and rabbit pellets ( Kaytee , Chilton , WI , USA ) was used for Culex sp . larvae . Adult mosquitoes were provided with 10% sucrose solution ad libitum for maintenance . For reproduction and virus infection purposes , adults were fed with expired anonymous human blood ( Biological Specialty Corporation , Colmar , PA , USA ) . Two strains of MAYV were used for the experimental infections: BeAr 505411 ( BEI Resources , Manassas , VA , USA ) , a genotype L strain isolated from Haemagogus janthinomys mosquitoes in Para , Brazil , in March 1991 , and BeAn 343102 ( BEI Resources , Manassas , VA , USA ) , a genotype D strain originally isolated from a monkey in Para , Brazil , in May 1978 . Both viruses were passed once in African green monkey kidney ( Vero ) cells . Virus-infected supernatant was aliquoted and stored at −70°C until used for mosquito infections . Viral stock titers were obtained by the focus forming unit ( FFU ) technique , as described below . Five- to seven-day-old females that had not previously blood-fed were used for experiments . Mosquitoes were allowed to feed on human blood spiked with virus via a glass feeder jacketed with 37°C distilled water for 1 h . Aliquots of the infectious bloodmeals were collected and titers of MAYV were determined by FFU ( Table 1 ) . After blood feeding , mosquitoes were anesthetized and fully engorged females were selected and placed in cardboard cages . Infection rate ( IR ) , dissemination rate ( DIR ) , transmission rate ( TR ) , and transmission efficiency ( TE ) were assessed at 7 and 14 days post-infection ( dpi ) . IR was measured as the rate of mosquitoes with infected bodies among the total number of analyzed mosquitoes . DIR was measured as the rate of mosquitoes with infected legs among the mosquitoes with positive bodies . TR was measured as the rate of mosquitoes with infectious saliva among the mosquitoes with positives legs , and TE measured as the rate of mosquitoes with infectious saliva among the total number of analyzed mosquitoes [19] . At 7 and 14 dpi , mosquitoes were anesthetized with triethylamine ( Sigma , St . Louis , MO , USA ) . Legs were detached from each body and placed in 2-mL tubes filled with 1 mL of mosquito diluent ( 20% heat-inactivated fetal bovine serum [FBS] in Dulbecco’s phosphate-buffered saline [PBS] , 50 μg/mL penicillin/streptomycin , 50 μg/mL gentamicin , and 2 . 5 μg/mL fungizone ) and a single zinc-plated , steel , 4 . 5-mm bead ( Daisy , Rogers , AR , USA ) , and tubes immediately placed on ice . Saliva was collected by forced salivation into a capillary tube as previously described [20] , expelled into in a 2-mL tube filled with 100 μL of mosquito diluent , and immediately placed on ice . Body and leg samples were homogenized at 30 Hz for 2 min using a TissueLyser II ( QIAGEN GmbH , Hilden , Germany ) and centrifuged for 30 sec at 11 , 000 rpm . All samples were stored at −70°C until tested . The presence of infectious MAYV particles in the body , legs , and saliva samples was tested by FFU assay in Vero cells . Vero cells were grown to a confluent monolayer in 96-well plates at 37°C with 5% CO2 in complete media ( Dulbecco’s modified-essential media [DMEM] , 100 units/mL penicillin/streptomycin , and 10% FBS ) . The next day , wells were washed with DMEM without FBS and incubated with a 30-μL of 10-fold serial dilutions ( 10−1 to 10−4 ) of each homogenized tissue sample for 2 h at 37°C . Saliva samples were not diluted . After the incubation step , the 30-μL aliquot was removed from the cell monolayer and any unattached viral particles removed with a DMEM wash . A total of 100 μL of overlay medium ( 1% methyl cellulose in complete growth medium ) was dispensed into each well , and plates incubated at 37°C with 5% CO2 . Cells were fixed at 24 h ( bodies and legs samples ) or 48 h ( saliva samples ) post-infection with 4% paraformaldehyde ( Sigma , St . Louis , MO , USA ) . Fixed cells were blocked and permeabilized for 30 min with blocking solution containing detergent ( 3% bovine serum albumin and 0 . 05% Tween 20 in PBS ) and washed with cold PBS . Viral antigens in infected cells were labeled using the monoclonal anti-chikungunya virus E2 envelope glycoprotein clone CHK-48 ( which reacts with Alphaviruses ) ( BEI Resources , Manassas , VA , USA ) diluted 1:500 in blocking solution . Subsequently , cells were washed 4 times with cold PBS to remove unbound primary antibodies . The primary antibody was labeled with the Alexa-488 goat anti-mouse IgG secondary antibody ( Invitrogen , Life Science , Eugene OR , USA ) at a dilution of 1:500 , and green fluorescenct foci observed and enumerated with an Olympus BX41 microscope equipped with an UPlanFI 4× objective and a FITC filter . Fluorescent foci were counted by eye ( the dilution with less than 100 foci was selected for each sample ) and virus titers calculated and expressed as FFU/mL . Data were analyzed using GraphPad Prism version 7 . 04 . Differences in the IR , DIR , TR , and TE of mosquitoes challenged with BeAr 505411 and BeAn 343102 were analyzed by Fisher’s exact test . The Mann-Whitney U test was used to compare the body , legs , and saliva viral titers of mosquitoes exposed to BeAr 505411 or BeAn 343102 .
A total of 115 Ae . aegypti , 19 An . freeborni , 31 An . gambiae , 29 An . quadrimaculatus , 132 An . stephensi and 60 Cx . quinquefasciatus were analyzed in this study . Details of analyzed mosquitoes and the IR , DIR , TR , and TE are in Table 1 . All six mosquito species were susceptible to infection with MAYV to some degree , although there were MAYV strain–specific differences . IRs for Ae . aegypti exposed to strain BeAr 505411 were significantly higher compared to strain BeAn 343102 ( p<0 . 0001 ) at 7 dpi , and IRs for strain BeAr 505411 at 7 dpi were significantly higher than 14 dpi ( p<0 . 0001 ) . Moreover , no Ae . aegypti exposed to strain BeAn 343102 became infected at 14 dpi despite the presence of positive mosquitoes at 7 dpi . IRs for An . gambiae and An . stephensi were similar across MAYV strains , and IRs increased over time in An . gambiae . An . freeborni and An . quadrimaculatus were susceptible to infection with both strains of MAYV , and Cx . quinquefasciatus was susceptible only to a low-frequency infection with strain BeAr 505411 and refractory to BeAn 343102 infection . Once infected , all tested mosquito species developed a disseminated infection . Disseminated infection was generally detected as early as 7 dpi , with the exception of An . freeborni exposed to the BeAr 505411 strain . DIRs were similar for both virus strains in An . gambiae and An . stephensi at both timepoints and for Ae . aegypti at 7 dpi . There was a trend toward higher DIRs for strain BeAn 343102 compared to strain BeAr 505411 in An . quadrimaculatus and An . freeborni at day 7 . There was also a trend toward a higher DIR at 14 dpi than at 7 dpi for strain BeAr 505411 in Ae . aegypti , both strains in An . gambiae , and strain BeAr 505411 in An . freeborni . Transmission was detected in all Anopheles species and Ae . aegypti ( albeit very poorly ) , but not in Cx . quinquefasciatus . An . gambiae , An . quadrimaculatus and An . stephensi were able to transmit both MAYV strains tested . For Ae . aegypti only a single transmission event was detected for virus strain BeAr 505411 . Only virus strain BeAn 343102 was transmitted by An . freeborni . Both virus strains could be transmitted by An . gambiae , quadrimaculatus and stephensi . MAYV titers for all samples were calculated and expressed as FFU/mL . Ae . aegypti exposed to strain BeAr 505411 had significantly greater titers in the bodies ( 7 and 14 dpi ) and legs ( 7 and 14 dpi ) compared to strain BeAn 343102 ( p<0 . 0001 ) ( Fig 1 ) . Conversely , An . stephensi exposed to strain BeAn 343102 had significantly greater titers in the bodies ( 7 dpi , p<0 . 05; 14 dpi , p<0 . 001 ) and legs ( 7 dpi , p<0 . 001 ) compared to strain BeAr 505411 ( Fig 1 ) . There were no significant differences in body , legs , or saliva titers between the MAYV strains in An . freeborni , An . gambiae , An . quadrimaculatus and Cx . quinquefasciatus .
It should be noted that all results were conducted with laboratory mosquito strains . Results with other laboratory strains or wild mosquitoes that differ in their nuclear genotype or microbiome may differ from those we obtained . All mosquitoes in our experiments were infected with approximately 7 logs of virus; for mosquitoes that did not readily become infected it is possible that higher viral titers might result in higher infection rates . In our experiments we only tested two timepoints post-infection ( seven and fourteen days ) . It is possible that longer incubation times may result in different infection and/or transmission rates .
|
Mayaro virus ( MAYV ) is a mosquito-borne Alphavirus responsible for outbreaks in South America and the Caribbean . In this study we infected different species of mosquito ( belonging to the genera Aedes , Anopheles and Culex ) with MAYV and tested their capacity to transmit the virus at different time points . Results show that Anopheles mosquitoes were competent vectors for 2 genotypes of MAYV , while Aedes and Culex were poor vectors . The capacity of Anopheles mosquitoes to transmit MAYV highlights their importance as neglected vectors of arboviruses . These data suggest that Anopheles mosquitoes have the potential to sustain transmission cycles of neglected pathogens in naïve regions , including the United States .
|
[
"Abstract",
"Introduction",
"Material",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"species",
"colonization",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"viral",
"transmission",
"and",
"infection",
"invasive",
"species",
"pathogens",
"microbiology",
"saliva",
"animals",
"alphaviruses",
"viral",
"vectors",
"viruses",
"rna",
"viruses",
"insect",
"vectors",
"infectious",
"diseases",
"anopheles",
"gambiae",
"aedes",
"aegypti",
"medical",
"microbiology",
"culex",
"quinquefasciatus",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"arthropoda",
"mosquitoes",
"eukaryota",
"anatomy",
"mayaro",
"virus",
"physiology",
"viral",
"pathogens",
"virology",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"organisms"
] |
2018
|
Anopheles mosquitoes may drive invasion and transmission of Mayaro virus across geographically diverse regions
|
The cerebral cortex of mammals exhibits intricate interareal wiring . Moreover , mammalian cortices differ vastly in size , cytological composition , and phylogenetic distance . Given such complexity and pronounced species differences , it is a considerable challenge to decipher organizational principles of mammalian connectomes . Here , we demonstrate species-specific and species-general unifying principles linking the physical , cytological , and connectional dimensions of architecture in the mouse , cat , marmoset , and macaque monkey . The existence of connections is related to the cytology of cortical areas , in addition to the role of physical distance , but this relation is attenuated in mice and marmoset monkeys . The cytoarchitectonic cortical gradients , and not the rostrocaudal axis of the cortex , are closely linked to the laminar origin of connections , a principle that allows the extrapolation of this connectional feature to humans . Lastly , a network core , with a central role under different modes of network communication , characterizes all cortical connectomes . We observe a displacement of the network core in mammals , with a shift of the core of cats and macaque monkeys toward the less neuronally dense areas of the cerebral cortex . This displacement has functional ramifications but also entails a potential increased degree of vulnerability to pathology . In sum , our results sketch out a blueprint of mammalian connectomes consisting of species-specific and species-general links between the connectional , physical , and cytological dimensions of the cerebral cortex , possibly reflecting variations and persistence of evolutionarily conserved mechanisms and cellular phenomena . Our framework elucidates organizational principles that encompass but also extend beyond the wiring economy principle imposed by the physical embedding of the cerebral cortex .
Mapping and understanding the wiring of the cerebral cortex at the micro- , meso- , and macroscale level is a central challenge in neuroscience [1–9] . Extensive studies have mapped the structural connections among cortical areas—that is , the macroscale connectional architecture—in different mammals , such as cats [3] , mice [6 , 7] , and macaque and marmoset monkeys [8 , 10] . These studies have uncovered a characteristic pattern of cortico-cortical connections among cortical areas , providing the structural scaffold for the communication of cortical areas , which is essential for cognition and behavior [11–13] . Moreover , invasive tract-tracing studies in mammals have uncovered a graded variation in the laminar origin of connections [14–16] , a connectional feature related to physiological properties of long-range connections [12] , central to contemporary theories of brain structure and function [17 , 18] and the basis of the so-called hierarchical arrangement of the areas of the cerebral cortex [19] . From a network topology standpoint—that is , the arrangement of connections between the distinct areas of the cortex—cortical connectomes possess a tightly interconnected structural core [20 , 21] , a network topology that is considered important for flexible behavior and large-scale functional integration [22] . Moreover , mammalian species exhibit a divergent evolutionary history of millions of years , as well as pronounced differences with respect to , for instance , brain size , number of neurons , and duration of neurogenesis [23–28] ( Fig 1 ) . Given the intricate wiring configuration of the cortex and such pronounced differences , is it possible to decipher unifying principles that link the connectional architecture with other dimensions of cortical architecture and thus sketch out a blueprint of the cortical organization of mammals ? A principle related to the existence of connections is the wiring cost principle . Specifically , nearby areas are more likely to be connected than remote areas [21 , 29–31] . However , wiring cost , reflected in the physical distance between cortical areas , does not fully explain the existence of connections [21 , 32 , 33] . Qualitative observations in the macaque monkey cortex suggest that the existence of connections is closely related to the gradients of cytoarchitectonic differentiation of the cerebral cortex , specifically to the similarity of the degree of cytoarchitectonic differentiation of cortical areas [34] . Gradients of cytoarchitectonic differentiation are formed by spatially ordered changes in the cytological composition of areas , including the appearance of the granular layer ( layer IV ) and the increase of its neuronal density and width , as well as the successive distinguishability and increase of the neuronal density of upper ( supragranular ) layers compared to lower ( infragranular ) layers [16 , 27 , 35–38] . Systematic studies in different mammalian species have demonstrated that the similarity of cytoarchitectonic differentiation of cortical areas , above and beyond their physical distance , is closely related to the existence of connections , suggesting a common wiring principle of mammalian cortices [30 , 31 , 33 , 39] . It is , however , unknown if this wiring principle is manifested in a species-specific manner and how general it is across the mammalian phylogeny . With respect to the systematic shifts of the laminar origin of connections , two main explanations have been put forward . On the one hand , a framework postulates that the graded shift of the laminar origin of connections , from predominantly infragranular to predominantly supragranular , giving rise to “feedback” and “feedforward” type of connections , respectively , is manifested across the rostrocaudal axis of the brain [14 , 40–42] . On the other hand , a cytoarchitecture-based framework has emphasized the central role of cortical cytoarchitectonic gradients in relation to the gradual shifts of the laminar origin of connections [15 , 30 , 33 , 39 , 43] . Therefore , a conjoint examination of these alternative frameworks in different mammalian cortices is needed for deciphering the central dimension of cortical organization that is related to the graded shifts of the laminar origin of connections . From a network topology standpoint—that is , the arrangement of connections between the different areas of the cortex—a core–periphery structure characterizes the mouse and macaque monkey cortex [20 , 21] . The core–periphery structure corresponds to two sets of areas , a tightly interconnected set of areas constituting the core and the rest of the areas constituting the periphery . This network configuration is central to theories of animal cognition [22] . In the macaque monkey , the core–periphery division is also reflected in the cytoarchitecture of the cortex , thus offering a unifying principle linking network topology and cytology by elucidating the cellular composition of topologically central cortical areas and the potential neurodevelopmental mechanisms leading to their central role in the cortical connectome [33] . Therefore , it is important to elucidate the species-general or species-specific nature of the relation of the core–periphery structure to the cytology of the cortex across different mammals . Here , we examine the connectomes of the mouse , cat , and macaque and marmoset monkey . We relate the connectional , cytoarchitectonic , and physical dimensions of the cerebral cortex , thus highlighting unifying principles that link the different dimensions of cortical architecture . These principles are manifested in a species-general but also species-specific manner , distinguishing the mouse and the marmoset monkey from the cat and the macaque monkey cortex . Commonalities allow the extrapolation of connectional features to unexamined species , such as humans , whereas the species-specific principles point at potential functional differences across species and indicate varied degrees of vulnerability to pathology . The observed unifying principles may reflect variations of evolutionary conserved neurodevelopmental mechanisms .
Extensive cytoarchitectonic and connectome data as well as information on the physical distance between cortical areas of mouse , cat , and marmoset and macaque monkey cortices were used in the analyses [8 , 10 , 30 , 31 , 33 , 44 , 45] . For the mouse and cat cortex , the cytoarchitectonic differentiation of areas was assessed qualitatively by defining an ordinal scale of cortical types based on Nissl-stained sections [30 , 31] . Cortical types reflect a multidimensional characterization of the cytoarchitectonic differentiation of cortical areas , based primarily on the density of neurons in the different cortical layers , as well as the appearance , neuronal density , and thickness of layer IV [39] . Low cortical types—that is , less differentiated and overall less neuronally dense areas—are not clearly laminated , and layer IV is absent or only weakly present . By contrast , high cortical types—that is , more differentiated and overall more neuronally dense areas—are clearly laminated , with a clearly defined layer IV . By these criteria , the highest cortical type corresponds to areas such as the primary visual cortex [30 , 39] . Therefore , cortical areas constitute a cortical spectrum of cytoarchitectonic differentiation , ranging from less to more differentiated and , thus , overall neuronally dense cortical areas ( Fig 1 ) . The cytoarchitectonic differentiation of the marmoset and macaque monkey cortical areas was assessed quantitatively by their overall neuronal density—that is , the number of neurons per mm3 . Neuronal density constitutes a fingerprint of the cytoarchitectonic status of cortical areas [33 , 36 , 37] . For the macaque monkey , Nissl- and NeuN-stained material was used [33] . For the marmoset monkey , NeuN-stained material was used [44] . Qualitative assessment of cytoarchitectonic differentiation of the areas of the macaque monkey , based on cortical types , was used as a control analysis . We used the most-comprehensive available cortical connectomes of the mouse [7 , 21] , cat [3] , and marmoset [10] and macaque monkey [8] . We used the geodesic or Euclidean distance between the barycenters of the cortical areas as a measure of their physical distance [8 , 10 , 21 , 31] . In the absence of a stereotaxic atlas with the parcellation scheme of Scannell and colleagues [3] for the cat cortex , physical distance between areas of the cat cortex was defined as the number of areas separating a pair of areas [30] . The presence or absence of a connection was viewed against two dimensions of cortical organization—that is , the cytoarchitectonic and physical dimension ( Fig 2 ) . We verified that connections that are present span shorter distances than absent connections ( statistical energy test: 0 . 17 , 0 . 74 , 0 . 03 , 0 . 33 for the mouse , cat , and marmoset and macaque monkey , respectively; all p < 0 . 001 ) . Moreover , connections that are present involve pairs of areas with more-similar cytoarchitecture than areas that are not connected ( statistical energy test: 0 . 32 , 0 . 29 , 0 . 16 , 0 . 23 for the mouse , cat , and marmoset and macaque monkey , respectively; all p < 0 . 001 ) . Similarity of cytoarchitecture was assessed as the absolute difference of the cortical type or neuronal density of a pair of areas . Fig 2 summarizes these findings , demonstrating that cytoarchitectonic similarity of cortical areas and their physical distance relates to the existence of connections . Using an alternative dataset for mouse connectivity [21] and qualitative assessment of the cytoarchitectonic status of the areas for the macaque monkey cortex led to similar qualitative results ( S1 Fig ) . Conjoint examination of the role of cytoarchitectonic similarity and physical distance to existence of connections with multivariate logistic regression revealed a statistically significant contribution of both factors in all species with the exception of the marmoset monkey , in which cytoarchitectonic similarity did not reach statistical significance ( S2 Fig ) . Thus , in the marmoset monkey , cytoarchitectonic similarity does not relate , above and beyond physical distance , to the pattern of existence of connections among cortical areas . This discrepancy constitutes the first species-specific manifestation of the relation of cytoarchitecture and connectivity . Control analyses for the species for which an ordinal scale was used in assessing the cytoarchitectonic status of cortical areas—that is , the cat and mouse—revealed that the relation of cytoarchitectonic similarity and existence of connections was robust to the exact assignments of cortical types to areas , as well as the exact range of the ordinal scale used for the qualitative evaluation of cytoarchitectonic differentiation in these species ( S3 Fig ) . To further investigate species-specific relations of cytoarchitecture and connectivity , we performed a logistic regression analysis for the mammals that showed a significant relation between cytoarchitecture and connectivity in the multivariate logistic regression analysis—that is , the mouse , cat , and macaque monkey . For each pair of these mammals , a model was estimated with existence of connections as the binary dependent variable and cytoarchitectonic similarity and distance as regressors . For investigating species-specific effects , a further regressor coding for the different species and their interaction with cytoarchitectonic similarity was added . Coefficients from the logistic regression denote the impact of each regressor on the probability of finding a connection between a pair of areas , as well as the dependence of such an effect on the interaction of the regressors . It should be noted that our approach does not require the establishment of area homologies across species , since the cross-species analysis relies on pairs of cortical areas and examines the factors related to the presence or absence of a connection between each pair of areas , irrespective of potential homologies or absence thereof . These analyses showed that the effect of cytoarchitectonic similarity on the existence of connections was significant in all cases , but its role was different when the mouse was compared with the cat and macaque monkey . For the mouse versus macaque monkey analysis , the coefficients were distance = −2 . 74 , cytoarchitectonic similarity = −0 . 80 , and species by cytoarchitectonic similarity = −2 . 47 ( all p < 0 . 0001 ) . The inclusion of the species by cytoarchitectonic similarity interaction significantly improved the model fit , as indicated by a likelihood ratio ( LR ) test ( LR = 16 . 95 , p < 0 . 001 ) . For the mouse versus cat analysis , the coefficients were distance = −1 . 67 , cytoarchitectonic similarity = −0 . 90 , and species by cytoarchitectonic similarity = −1 . 77 ( all p < 0 . 0001 ) . The inclusion of the species by cytoarchitectonic similarity interaction significantly improved the model fit , as indicated by an LR test ( LR = 20 . 55 , p < 0 . 001 ) . For the cat versus macaque monkey analysis , the coefficients were distance = −2 . 70 and cytoarchitectonic similarity = −2 . 76 , ( both p < 0 . 001 ) , but the interaction between the species and cytoarchitectonic similarity regressors was not significant ( species by cytoarchitectonic similarity = −0 . 53 , p > 0 . 1 ) . The negative coefficients for the interaction between species and cytoarchitectonic similarity for the mouse–cat and mouse–macaque monkey analyses were significant . This indicates that the impact of the decrease of the cytoarchitectonic similarity on the decrease of the probability of the existence of a connection is higher in the cat and macaque monkey when compared to the mouse ( see also the Materials and Methods section ) . For a better understanding of this effect , we visualized the impact of cytoarchitectonic similarity on the probability of the existence of a connection for different physical distance values for all pair-wise species analyses ( Fig 3 ) . For an equal decrease of cytoarchitectonic similarity , the probability of the existence of a connection decreased more slowly for the mouse when compared to the cat and macaque monkey ( Fig 3 ) . A control analysis using the qualitative cytoarchitectonic status of the macaque monkey cortical areas and a different dataset for the mouse cortico-cortical connectivity [21] led to the same qualitative results . Specifically , the coefficients for the mouse versus cat analysis were distance = −2 . 21 , cytoarchitectonic similarity = −0 . 84 , and species by cytoarchitectonic similarity = −1 . 87 ( all p < 0 . 001 ) . The inclusion of the interaction of species and cytoarchitectonic similarity significantly improved the model fit , as indicated by an LR test ( LR = 19 . 49 , p < 0 . 001 ) . For the mouse versus macaque monkey analysis , the coefficients were distance = −3 . 21 , cytoarchitectonic similarity = −0 . 77 , and species by cytoarchitectonic similarity = −2 . 07 ( all p < 0 . 01 ) . Also in this case , the inclusion of the interaction of species by cytoarchitectonic similarity significantly improved the model fit , as indicated by an LR test ( LR = 19 . 23 , p < 0 . 001 ) . Finally , for the cat versus macaque monkey analysis , the coefficients were distance = −3 . 27 and cytoarchitectonic similarity = −2 . 83 ( both p < 0 . 001 ) , but the interaction of the species and cytoarchitectonic similarity regressors was not significant ( species by cytoarchitectonic similarity = 0 . 01 , p > 0 . 1 ) . A visual depiction of the different effect of cytoarchitectonic similarity on the existence of connections in the mouse compared to the cat and macaque monkey for this control analysis is provided in ( S4 Fig ) . In summary , cytoarchitectonic similarity relates to the existence of connections in mammalian cortices in a species-specific manner , differentiating the mouse and marmoset monkey from the cat and macaque monkey . Next , we aimed at deciphering the central dimension of cortical organization related to the graded shifts of the laminar origin of connections across the different cortical areas . This analysis focused on the cat and macaque monkey , for which quantitative data on the laminar origin of connections were available [45 , 46] . A cytoarchitecture-based model was used , based on evidence from the prefrontal [15] and visual cortex [33 , 39] of the macaque monkey , highlighting the cytoarchitectonic status of the interconnected areas as predictive of the laminar origin of the connections . Here , we used quantitative information—that is , the percentage of supragranular labeled neurons ( NSG% ) [45] ( Fig 4A ) —that extends beyond the visual system of the macaque monkey and encompasses the rest of the cortex . We examined the cytoarchitecture-based model [15 , 43] conjointly with a rostrocaudal-based model that corresponds to suggestions that the rostrocaudal axis of the cortex is a central predictive factor of the laminar origin of connections [14 , 40–42] . We used support vector regression and partial Spearman's rank correlations . The conjoint examination of the relation of the laminar origin of connections to the rostrocaudal axis and cytoarchitectonic gradients was also performed with data from the cat cortex ( see Materials and Methods ) . In the macaque monkey , the cytoarchitecture-based model explained significantly more variance of the NSG% values than the rostrocaudal-based model ( Fig 4B and 4C ) . The addition of the rostrocaudal distances as a predictor to the cytoarchitecture-based model did not lead to statistically better NSG% predictions , compared to the model based solely on the cytoarchitecture of areas ( p > 0 . 1 ) ( Fig 4B and 4C ) . Specifically , the Spearman's rank correlation between the actual and predicted NSG% values for the cytoarchitecture-based model was rho = 0 . 36 , for the rostrocaudal-based model rho = 0 . 26 , and for the combination of cytoarchitecture and rostrocaudal distances rho = 0 . 37 ( all p < 0 . 0001 ) . The cytoarchitecture-based model explained more variance than the rostrocaudal-based model when partial Spearman's rank correlations were estimated: the correlation between NSG% and cytoarchitecture , when partialing out the rostrocaudal distances , was rho = 0 . 27 ( p < 0 . 0001 ) , and the correlation between NSG% and rostrocaudal distances when partialing out the cytoarchitectonic status of cortical areas was rho = 0 . 10 ( p < 0 . 05 ) . The same qualitative results were obtained when using the qualitative scale for assessing the cytoarchitecture of the macaque monkey cortex ( S5 Fig ) . Moreover , when computing partial Spearman's rank correlations for this control analysis , the cytoarchitecture-based model explained more variance than the rostrocaudal-based model: the correlation between NSG% and cytoarchitecture when partialing out the rostrocaudal distances was rho = 0 . 45 ( p < 0 . 0001 ) , and the correlation between NSG% and rostrocaudal distances when partialing out the cytoarchitectonic status of cortical areas was rho = 0 . 08 ( p < 0 . 05 ) . Moreover , recent studies advocating the importance of the rostrocaudal axis in predicting the laminar origin of connections in the macaque monkey cortex exclude the less differentiated areas of the cingulate and insular cortex [42] . Thus , we also performed the NSG% predictions while excluding these less differentiated cortical areas . This control analysis led to the same qualitative results—that is , the cytoarchitecture-based model yielded the highest NSG% predictions—and the addition of the rostrocaudal distances did not carry any additional information ( S6 Fig ) . The cytoarchitecture-based model also resulted in the best predictions of the NSG% values for the cat cortex ( Fig 5A ) . The same pattern of results as for the macaque monkey cortex was observed; namely , the cytoarchitecture-based model explained significantly more variance of NSG% values when compared to the rostrocaudal-based model ( Fig 5B and 5C ) . The addition of the rostrocaudal distances as an extra predictor to the cytoarchitecture-based model did not lead to statistically better NSG% predictions compared to the model based solely on the cytoarchitecture of areas ( p > 0 . 1 ) ( Fig 5B and 5C ) . Specifically , the Spearman's rank correlation between the actual and predicted NSG% values for the cytoarchitecture-based model was rho = 0 . 80 , for the rostrocaudal-based model rho = 0 . 21 , and for the combination of cytoarchitecture and rostrocaudal distances rho = 0 . 79 . The same conclusions were obtained for the partial Spearman's rank correlations: the correlation between NSG% and cytoarchitecture when partialing out the rostrocaudal distances was rho = 0 . 79 ( p < 0 . 0001 ) , and the correlation between NSG% and rostrocaudal distances when partialing out the cytoarchitectonic status of cortical areas was rho = 0 . 05 ( p > 0 . 1 ) . These conclusions are further supported by an analysis of an additional dataset with categorical data on the laminar patterns of the connections in the cat cortex ( S7 Fig ) . Our results highlight the cytoarchitectonic gradients of the cerebral cortex as a central axis of organization related to the graded shifts in the laminar origin of connections across the cortical sheet . The implications of these findings are 2-fold . First , they offer a guiding thread for deciphering the cellular phenomena or mechanisms responsible for such a close systematic relation between cytoarchitecture and laminar origin of connections ( see Discussion ) . Second , a cytoarchitecture-based model built on macaque monkey data can predict the laminar origin of connections in the human cortex , since such connectional data cannot currently be obtained by in vivo experiments ( Fig 6 ) . Such extrapolation of connectional features renders possible novel structure–function relations to be examined at a whole-cortex level , e . g . , relating interareal functional communication of cortical areas and the underlying laminar origin of the connections between them [47] , without the necessity of establishing macaque–human cortical area homologies . It is known that the mouse and macaque monkey cortex possesses nonrandom topological connectivity features , such as a core–periphery structure [20 , 21] . A core is a set of areas that are highly interconnected , with the remaining noncore areas constituting the periphery . A core–periphery network topology characterizes not only cortico-cortical networks but also other biological and technological networks , providing properties such as high topological efficiency [22 , 49] . We aimed to map this topological structure and investigate its association with the cytoarchitectonic gradients of the cerebral cortex of the different mammals . The core–periphery structure has already been mapped in the mouse and macaque monkey cortex by uncovering the largest cliques of the cortical connectome and forming the core as the union of areas participating in these cliques [20 , 21] ( Fig 7A and 7D ) . Here , to enable a comparative examination , we mapped the core–periphery structure with the same method in the cat and marmoset monkey . We found that the cat exhibits a core that consists of the union of two cliques of size 9 ( S3 Table ) , whereas the marmoset monkey core is composed of the union of 12 cliques of size 16 ( S4 Table ) . For both the cat and marmoset monkey , the size of the largest cliques forming the core was significantly different from the size of the largest cliques observed in random networks matched for degree distribution , number of nodes , and number of edges ( p < 0 . 001 for both the cat and marmoset monkey , 1 , 000 null networks ) . The cat core consists of areas that have a visuomotor and multisensory integration functional signature [3] ( Fig 7C ) . The marmoset core consists of “association” and multimodal areas of the frontal , parietal , and temporal lobe ( Fig 7B ) . We next related the core–periphery topology with the cytoarchitecture of the cerebral cortex . These analyses distinguished the mouse and marmoset monkey from the cat and macaque monkey . Specifically , in the mouse and marmoset monkey , the core areas did not significantly differ from the periphery areas in terms of cytoarchitectonic differentiation ( Fig 7A and 7B ) . On the contrary , in the cat and macaque monkey cortex , significant differences were observed between the core and periphery areas , with core areas exhibiting lower neuronal densities and degree of cytoarchitectonic differentiation than the periphery areas ( Fig 7C and 7D ) . Thus , although a structural network core characterizes the cortico-cortical network of all examined species , this topological structure is related differently to the cytology of the cerebral cortex , with a shift of the network core to less neuronally dense and differentiated areas of the cortex as the arbiter between the examined species . We subsequently proceeded to the explicit elucidation of the role of the network core in the communication among cortical areas . To this end , we examined the efficiency of the core areas , under two diametrically opposite and recently suggested scenarios of network communication [50] . Specifically , we assessed if the core exhibited higher efficiency under the scenario that communication in the cortical network takes place via the shortest paths or as passive diffusion ( corresponding to random walks ) [50] . For all species and both modes of communication ( shortest path or random walk ) , core areas exhibited higher incoming efficiency than the periphery areas ( Fig 8 ) . Thus , the core , compared to the periphery , can be reached faster from cortical areas under both modes of communication . Moreover , for all species , the core areas also exhibited higher outgoing efficiency for shortest paths but not for the random walk mode of communication ( Fig 8 ) . Thus , under the random walk mode of communication , core areas , compared to the periphery , are not topologically privileged for fast access to other cortical areas . Hence , in the context of the two aforementioned modes of network communication , the structural core must adhere to a mode of communication geared toward shortest paths in order to achieve fast access to the areas of the cortical network . In sum , the structural core , which is central for network communication , constitutes a common network topology of diverse mammals . The core–periphery topology is related to the cytology of the cortex in a species-specific manner . Specifically , a displacement takes place toward the less differentiated and overall neuronally dense areas of the cerebral cortex when transitioning from the mouse and marmoset monkey to the cat and macaque monkey .
The premise that neuronal systems are wired in such a way that minimizes the physical distance between the interconnected elements explains part of the characteristic pattern of presence and absence of connections between cortical areas in different mammalian species [9 , 20 , 21 , 29–31] . The current examination demonstrated that wiring cost also constrains the cortical connectome of a New World monkey—that is , the marmoset monkey . The current comparative framework allowed us to gain deeper insights into how the relation of cytoarchitectonic similarity to the existence of connections manifests across species . Our results reveal that cytoarchitectonic similarity has an attenuated impact on the probability of the existence of connections in the mouse when compared to the cat and macaque monkey and is statistically absent in the marmoset monkey . Thus , unifying wiring principles linking connectivity and cytoarchitecture not only distinguish rodents from primates but also point out differences within the primate order . Primates are distinguished from rodents with respect to the scaling of the size of the brain and the number of neurons it contains [55] . Our study offers further comparative insights by relating the cytology of the cortex to its macroscale connectivity and assessing how this relation is manifested in different mammals . The species-specific manifestation of the relation of connectivity and cytoarchitecture is systematic and highlights the trajectory of this relation across the mammalian spectrum . Specifically , our framework predicts that , on average , the relation of cytoarchitectonic similarity and the existence of connections in smaller mammalian cortices ( e . g . , hamster , treeshrew ) might be attenuated or even absent when compared to larger mammalian cortices ( e . g . , great apes , humans ) . Modifications of evolutionarily conserved developmental mechanisms may be the cause for this species-specific relation of cytology and connectivity ( Fig 9 ) ( see “Heterochronous , graded neurogenesis and pyramidal cell size heterogeneity” ) . The current results complement and resonate well with recent findings that show a common , but also species-specific manifestation , of the role of physical distance in the wiring of the mouse and macaque monkey cortex [21] . The current quantitative cross-species examination also allows to decipher the central cortical dimension that relates to the graded shifts of the laminar origin of cortico-cortical connections . In both the cat and macaque monkey , the laminar origin of the connections is dictated by the cytoarchitectonic status of the interconnected areas and not the orientation of a connection along the rostrocaudal spatial axis . Thus , the current investigation , conjointly with previous results [33 , 39 , 56] , highlight the close relation of cortical cytoarchitectonic gradients to the systematic shifts of the laminar origin of connections . The demonstration that cytoarchitectonic differentiation is a central cortical dimension related to the laminar origin of connections offers the ground for deciphering the concrete cellular phenomena across the cortical sheet that might be responsible for the shifts of the laminar origin of cortico-cortical connections ( see “Heterochronous , graded neurogenesis and pyramidal cell size heterogeneity” ) . In addition , the close relation of cytoarchitectonic gradients and laminar origin of connections can be used for extrapolating this connectional feature to the human cortex ( see “Unifying wiring principles allow cross-species predictions” ) . We should also note that the cytoarchitectonic gradients of the cerebral cortex explain a substantial part of the variance of the shifts of the laminar origin of connections , but not the total variance; thus , it is important to uncover additional factors that may shape the laminar shifts of connections across the cortical sheet of mammals . The species-general finding that cytoarchitectonic gradients constitute a central cortical dimension related to the laminar origin of connections not only offers neurobiological insights but also allows the extrapolation of this connectional feature from macaque monkeys , the closest primate to humans that can be invasively examined , to the human cerebral cortex . Such extrapolation allows novel structure–function examinations . For instance , frequency-dependent communication between areas of the human visual system obeys the same structure–function principles observed in the macaque monkey [12 , 47] . Our results allow extrapolation of the laminar origin of connections to humans , thus allowing such structure–function examinations to be performed at the global , whole-cortex level without the need for establishing homologies between the cortical areas of the two species . In sum , our results , conjointly with recent efforts [57 , 58] , demonstrate the value of uncovering unifying principles that can be used to link a wealth of data on the macaque monkey cerebral cortex to the human cerebral cortex . Our results highlight a structural network core in the mammalian cerebral cortex , important for the communication of cortical areas . Our comparative analysis demonstrates the displacement of the structural network core in the mammalian phylogeny . This displacement is manifested as a species-specific relation of the network core to the cytology of the cortex , leading to the neuronal sparsification of the core in cats and macaque monkeys—that is , the displacement of the network core toward the least neuronally dense parts of the cerebral cortex . We have demonstrated that across mammalian species , the structural core , compared to the periphery , is reached faster from cortical areas under two modes of network communication—that is , communication based on passive diffusion or shortest paths [50] . Passive diffusion is not costly from an information point of view , since navigating the network relies on random transitions from area to area , with no “knowledge” about the topology of the network . Shortest paths , on the other hand , require the channeling of communication between cortical areas through the shortest routes of the cortical network , and thus this mode of communication is considered information-costly [50] . Passive diffusion traps signals inside the densely interconnected core; hence , the structural core can reach faster , compared to the periphery , other areas only under the adoption of a more information-costly mode of communication that is geared toward shortest paths . Topologically central parts of the brain also exhibit a high energetic cost [59 , 60] , and thus a substantial part of this energetic cost , or overall energy consumption of the mammalian brain , might be attributable to the need of the areas of the core to adopt an information-costly mode of communication in order to achieve fast communication with other cortical areas . In sum , we highlight the importance of the structural network core in the communication of cortical areas . These findings provide an empirical foundation to theoretical frameworks [22] and extend previous results [9] by situating the structural network core in a comparative context and elucidating its role in light of recently suggested taxonomies of network communication processes . The displacement of the network core across the cortical gradients , leading to its neuronal sparsification in cats and macaque monkeys , highlights three points with potential functional ramifications and suggests varied degrees of vulnerability to pathologies . First , less cytoarchitectonically differentiated—and thus overall less neuronally dense—areas in mammalian cortices , such as the areas of the insular and cingulate cortex , are also overall less myelinated when compared to more differentiated areas , such as primary sensory-motor areas [35 , 61–63] . Both in vivo and in vitro studies in mammals demonstrate that high degree of myelination suppresses synaptic plasticity and axonal growth [64–66] . Thus , less myelinated areas are more flexible than more myelinated areas [67] . Therefore , the cat and macaque monkey network core , in contradistinction to the core of the mouse and , to a certain extent , marmoset monkey , seems to include the most-flexible areas of the cerebral cortex , bestowing the network core in these species with higher degrees of adaptability . Second , in the macaque monkey , spine densities vary across cortical areas , with less differentiated and overall neuronally dense areas exhibiting high spine densities . In mice , differences of spine densities across cortical areas are very attenuated [68 , 69] . Computational modeling employing the heterogeneity of spine densities across areas as a proxy for the strength of excitatory input to pyramidal cells demonstrates that spine density heterogeneity is important for the generation of temporal receptive windows across cortical areas , bestowing the less differentiated parts of the cortex with more-prolonged time windows that are ideal for integration of signals over longer time periods [45] . Thus , the above computational evidence and interspecies differences with respect to spine density heterogeneity across the cortical gradients indicate that in macaque monkeys , contrary to mice , the alignment of the network core with areas exhibiting high spine densities constitutes a synergy of connectional and microcytological features that may enhance the functional integration capacity of the core in macaque monkeys . Third , although the neuronal sparsification of the network core might entail differences in terms of integration and plasticity , it might also entail an increased vulnerability . Less neuronally dense areas also exhibit high metabolism and cellular stress [67] . Highly connected areas , like the areas of the network core that we have currently highlighted , are more affected in diverse pathologies [70] . Thus , a cortex characterized by a synergy between highly connected and neuronally sparse areas can also entail an increased vulnerability to pathologies . In sum , our results highlight the relation of the network core to the cytology of the cortex across different mammals and the functional ramifications of such relation , thus situating the topology of the mammalian connectome in a comparative and neurobiologically interpretable context . Uncovering unifying wiring principles of the cerebral cortex harnesses the complexity of cortical wiring and renders possible a glimpse into the neurodevelopmental mechanisms suggested by these principles . The more pronounced relation of cytoarchitectonic similarity to the existence of connections observed in cats and macaque monkeys in relation to mice and marmoset monkeys may be rooted in the spatiotemporal structure of neurogenetic gradients during development [16 , 36 , 53 , 71 , 72] . Specifically , the spatially ordered cytoarchitecture of cortical areas might reflect the spatially ordered heterochronous neurogenesis and subsequent migration of neurons across the developing pallium . Hence , areas with similar cytoarchitecture might also exhibit similar developmental time courses [27 , 36 , 73] . Therefore , areas with similar cytoarchitecture in the adult cortex might be more likely to be connected , since during development they host neurons that constitute more-readily available connection partners , following a “what develops together , wires together” principle [39] . This mechanistic explanation assigns a central role to the heterochronicity of neurodevelopmental events in the formation of intricate wiring configurations . Such a mechanism is directly supported by empirical studies investigating the neurogenesis and connections of the olfactory bulb and the primary olfactory cortex in rats [72] and computational modeling [74 , 75] . The duration of neurogenesis is shorter in mice compared to macaque monkeys [76] and possibly arises from a common evolutionarily conserved mechanism [54] . Less distinct time windows and overall shorter neurogenesis in the mouse and the marmoset monkey , when compared to the cat and macaque monkey , may result in the currently observed species-specific manifestation of the relation of existence of connections and the cytological composition of cortical areas ( Fig 9 ) . In addition , computational modeling [77] and empirical evidence from Caenorhabditis elegans suggest that heterochronicity in neurogenesis might also partially explain the formation of a structural core; that is , neurons that constitute a tightly interconnected core in the adult worm are born earlier than noncore neurons [78] . A similar “early neurogenesis advantage” , in addition to spatial constraints imposed by the geometry of the cortex [20] or the distinct molecular signature of core areas [79] , might constitute factors that lead to the formation of a network core in the mammalian cortex . Our results demonstrate a tight relation of laminar origin of connections and cytoarchitecture . Why do less differentiated cortical areas elicit connections predominantly from infragranular layers , whereas more-differentiated areas elicit connections from predominantly supragranular layers ? Part of the answer might lie in the phenomenon of externopyramidization ( Externopyramidisierung ) [38 , 80 , 81] . This structural organization principle of the cerebral cortex , observed in diverse species , including humans [38 , 61 , 80 , 82] , describes the rate of change of the ratio of the soma size of pyramidal neurons located in upper ( supragranular ) layers versus lower ( infragranular ) layers across the cortical sheet . In mammalian cortices ( for instance , cat and monkey cortices ) , the progressive differentiation of areas is accompanied by an increase of the soma size of supragranular pyramidal cells relative to the soma size of the infragranular pyramidal cells [81] . A larger soma size of pyramidal cells entails larger axon diameters , higher conduction velocities , and larger boutons that contain a higher number of vesicles , leading to higher probabilities and larger amounts of neurotransmitter release [81] . Therefore , the soma size of pyramidal neurons entails ultrastructural and functional properties of the corresponding axons , possibly rendering connections originating from pyramidal neurons with large soma more suitable for high-throughput long-range communication . Consequently , the phenomenon of externopyramidization might partially explain why less differentiated areas can establish and maintain long-distant connections primarily from infragranular layers , whereas more differentiated areas primarily from supragranular layers [81] . The phenomenon of externopyramidization is manifested with a varied degree of prominence across the mammalian spectrum , thus allowing the prediction of the laminar origin of connections in not-yet-examined species [81] . Specifically , in species with an attenuated manifestation of the phenomenon of externopyramidization , like mice , less pronounced shifts of the laminar origin of connections across the cortical sheet will be observed , whereas in species with a more prominent manifestation of the phenomenon of externopyramidization , like gorillas and humans , more pronounced shifts of the laminar origin of connections will be observed , with a gradual emphasis on supragranular layers [81] . In sum , our results highlight specific neurogenetic and cellular phenomena giving rise to unifying principles linking the cytoarchitectonic and connectional organization of the adult mammalian cortex . Gradients of cortical differentiation entail changes of multiple cortical features , such as myelin and density of different receptors and interneuron subtypes [16 , 80 , 83] . Thus , apart from obtaining more comprehensive quantitative cytoarchitectonic data , future studies in mammals should also elucidate how macroscale connectivity relates to other dimensions of cortical architecture . Moreover , changes across cortical gradients are layer-specific [27] . Therefore , in order to reveal a more fine-grained picture of cortical architecture , quantitative measurements should ideally also be obtained in a layer-wise manner . Furthermore , additional features such as the strength heterogeneity of connections , as well as new results from invasive tract-tracing studies [84] , should be examined . In the macaque monkey , connectivity strength heterogeneity is related not only to the physical embedding of the cortex but also to the homophily principle—that is , the connectional similarity of cortical areas [85] . Thus , we predict that the homophily principle will help explain the strength heterogeneity of cortico-cortical connections in other mammals . Lastly , in phylogenetically close species , such as monkeys and humans , common long-range fiber systems can be discerned and used for the examination of species-general and species-specific organizational principles [86] . Such an approach , in conjunction with the approach that we have adopted , increases the tools for quantitative cross-species examinations and hopefully will further pave the way for additional insights into the organization of mammalian cortices . Our results sketch out a connectional blueprint for the mammalian cerebral cortex by demonstrating species-general and systematic species-specific unifying principles linking the connectional , cytological , and physical dimensions of the cerebral cortex . The common principles allow the extrapolation of connectional features to not-yet-examined mammalian species , whereas the species-specific variations highlight unique aspects of cortical organization across the mammalian spectrum with potential function ramifications . Commonalities and differences of cortical organization may stem from variations and persistence of evolutionarily conserved neurodevelopmental mechanisms and cellular phenomena .
We used binary logistic regression for the prediction of the existence of connections across species . In order to render cross-species predictions feasible , the predictors ( physical distance and cytoarchitectonic similarity ) were linearly normalized to the 0–1 interval separately for each species . Note that this normalization does not artificially expand or shrink the levels of differentiation or size of each species , since the relative changes indicated by these regressors are of importance . Subsequently , a model was built with the existence of connections as a binary dependent variable and the physical distance and cytoarchitectonic similarity as predictors . We were interested in investigating if the cytoarchitectonic similarity of cortical areas relates to the existence of connections in a species-specific manner . Therefore , a categorical predictor coding for the different species was added to the model as well as the interaction of this predictor and the cytoarchitectonic similarity predictor . The improvement of the model fit , when the interaction of species and cytoarchitectonic similarity was included , was assessed with the LR test . For predicting the laminar origin of connections , an out-of-sample classification approach was adopted . We used support vector regression with a regularization parameter C = 1 . A cytoarchitecture-based model , quantifying the difference of the cytoarchitecture of the cortical area of projection origin versus the cytoarchitecture of the cortical area of projection termination , was built on 70% of the data and tested on the remaining data . The predictions were computed 1 , 000 times , each time using 70% of the available data to build the model ( drawing without replacement ) . The quality of the predictions was assessed by computing the Spearman's rank correlation between actual and predicted NSG% values . In the same fashion , a rostrocaudal-based model was built and tested . The coordinates of the barycenters of the cortical areas along the rostrocaudal axis were normalized to the 0–1 interval , with 0 denoting the most caudal area and 1 the most rostral area . Subsequently , for each connection the rostrocaudal coordinate of the connection origin was subtracted from the rostrocaudal coordinate of the connection termination . Hence , increasingly positive ( negative ) values of this rostrocaudal distance metric denote increasing rostral-to-caudal ( caudal-to-rostral ) distances . A 3D stereotaxic atlas was used for the macaque monkey [8] . For the cat cortex , in the absence of a 3D stereotaxic atlas , we used the 2D atlas of Scannell and colleagues [3] . The map was digitally reproduced , each cortical area was color-coded with a unique color , and the map was imported in MATLAB . Each area was assigned to a position along the rostrocaudal axis in this native coordinate system by computing the mean coordinate in the y-axis of all the pixels belonging to each area , and subsequently rostrocaudal distances were computed as described for the macaque monkey cortex . For estimating the unique variance explained by each predictor , we additionally computed partial Spearman's rank correlations between the NSG% values and the rostrocaudal distance and the cytoarchitectonic difference of the connection origin and termination . For the cat cortex , the laminar origins of the connections were available either as quantitative NSG% values or as a binary category—that is , “feedforward” or “feedback . ” For the NSG% values , the Spearman's rank correlation between the predicted and actual NSG% values was used for assessing the quality of the predictions . For the binary case , the quality of the predictions was assessed by computing the corresponding area under the curve of the receiver operating characteristic curves . Null predictions and significance levels were obtained by training the model 100 times on shuffled NSG% values or binary labels . For detecting the core in the cortico-cortical network , we followed the approach described in [20] . We used a MATLAB implementation of the Bron-Kerbosch algorithm with pivoting and degeneracy ordering ( https://de . mathworks . com/matlabcentral/fileexchange/47524-find-maximal-cliques-for-large—sparse-network ) to detect the largest cliques ( that is , sets of fully connected areas ) . The core was defined as the union of areas participating in the largest cliques , and the rest of the areas were assigned to the periphery . Applying this algorithm to the macaque monkey and mouse data resulted in the exact same core areas as the ones reported in [20] and [21] . We applied the same algorithm for detecting the core areas of the cat and marmoset monkey connectome . To test if the core observed in the empirical networks was not solely the result of the degree distribution heterogeneity of the networks , we applied the core–periphery algorithm as described above to 1 , 000 surrogate networks , matched for degree distribution , nodes , and edges to the empirical networks . The statistical significance of the core was computed by examining if the size of the largest cliques ( constituting the core ) in the surrogate networks exceeded the size of the largest cliques in the empirical networks . We used the Brain Connectivity Toolbox ( https://sites . google . com/site/bctnet/ ) [91] for estimating the in- and out-efficiency ( based on shortest paths or random walks ) of cortical areas . Because of an absence of quantitative information on the strength of connections for the mouse and cat connectomes , these measures were computed in binary connectomes . We used permutation tests for comparing the in- and out-efficiency ( based on shortest paths or random walks ) and cytoarchitectonic differentiation of the core and periphery areas . The labels of the areas denoting if they belong to the core or the periphery were permuted , and the core–periphery differences were estimated with the Kolmogorov-Smirnov test or the statical energy test , a nonparametric test for comparing two distributions [92] ( https://github . com/brian-lau/multdist/blob/master/minentest . m ) . The procedure was repeated 1 , 000 times , and the obtained null values were compared to the values obtained with the original core–periphery assignments .
|
The cerebral cortex is wired in a highly intricate manner and exhibits striking differences across mammals—for instance , in overall size and number of neurons . Here , we uncover common , but also species-specific , principles that link the physical , cellular , and connectional architecture of mouse , cat , and monkey brains . Commonalities allow the extrapolation of features to further , unexamined species , such as humans , whereas the species-specific principles point at potential functional differences , but also varied degrees of vulnerability , of mammalian brains . The observed unifying principles may reflect variations of evolutionarily conserved neurodevelopmental mechanisms . In sum , we sketch out a blueprint of mammalian cortical organization that elucidates the links between the physical , cytological , and connectional architecture . Our results indicate that caution is warranted when translating findings from one mammalian species to another , since some , but not all , cortical organizational properties are common across the mammalian spectrum .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"nervous",
"system",
"neuronal",
"differentiation",
"brain",
"vertebrates",
"marmosets",
"neuroscience",
"animals",
"mammals",
"cell",
"differentiation",
"primates",
"animal",
"models",
"developmental",
"biology",
"brain",
"mapping",
"experimental",
"organism",
"systems",
"old",
"world",
"monkeys",
"research",
"and",
"analysis",
"methods",
"monkeys",
"animal",
"studies",
"new",
"world",
"monkeys",
"animal",
"cells",
"macaque",
"connectomics",
"cellular",
"neuroscience",
"eukaryota",
"neuroanatomy",
"anatomy",
"cats",
"cell",
"biology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"amniotes",
"cerebral",
"cortex",
"organisms"
] |
2019
|
A blueprint of mammalian cortical connectomes
|
Various bacterial toxins circumvent host defenses through overproduction of cAMP . In a previous study , we showed that edema factor ( EF ) , an adenylate cyclase from Bacillus anthracis , disrupts endocytic recycling mediated by the small GTPase Rab11 . As a result , cargo proteins such as cadherins fail to reach inter-cellular junctions . In the present study , we provide further mechanistic dissection of Rab11 inhibition by EF using a combination of Drosophila and mammalian systems . EF blocks Rab11 trafficking after the GTP-loading step , preventing a constitutively active form of Rab11 from delivering cargo vesicles to the plasma membrane . Both of the primary cAMP effector pathways -PKA and Epac/Rap1- contribute to inhibition of Rab11-mediated trafficking , but act at distinct steps of the delivery process . PKA acts early , preventing Rab11 from associating with its effectors Rip11 and Sec15 . In contrast , Epac functions subsequently via the small GTPase Rap1 to block fusion of recycling endosomes with the plasma membrane , and appears to be the primary effector of EF toxicity in this process . Similarly , experiments conducted in mammalian systems reveal that Epac , but not PKA , mediates the activity of EF both in cell culture and in vivo . The small GTPase Arf6 , which initiates endocytic retrieval of cell adhesion components , also contributes to junctional homeostasis by counteracting Rab11-dependent delivery of cargo proteins at sites of cell-cell contact . These studies have potentially significant practical implications , since chemical inhibition of either Arf6 or Epac blocks the effect of EF in cell culture and in vivo , opening new potential therapeutic avenues for treating symptoms caused by cAMP-inducing toxins or related barrier-disrupting pathologies .
Bacterial pathogens enhance infectivity by secreting toxins that deregulate immune signaling pathways or disrupt host cellular barriers . One class of toxins produced by diverse bacterial species dramatically increases intracellular concentrations of cAMP . This striking evolutionary convergence suggests that over-production of this second messenger represents a successful strategy to promote growth and dissemination of infectious agents and associated disease symptoms [1] . These toxins include adenylate cyclases ( AC ) , such as edema factor ( EF ) from Bacillus anthracis ( B . a . ) , CyaA from Bordetella pertussis , and ExoY from Pseudomonas aeruginosa . Other toxins modify host proteins to induce cAMP production by endogenous cellular machineries . For example , cholera toxin ( Ctx ) from Vibrio cholerae , and the related heat-labile toxin from enterotoxigenic Escherichia coli , both ADP-ribosylate the α subunit of trimeric G proteins to stimulate cAMP synthesis by host AC , while pertussis toxin ( Ptx ) from Bordetella pertussis ADP-ribosylates and inactivates Gi subunits that normally inhibit endogenous ACs ( reviewed in [2] ) . B . a . , the etiological agent of anthrax , produces two A-subunit toxins , edema factor ( EF ) and lethal factor ( LF ) , which are secreted together with a shared B-subunit , protective antigen ( PA ) , and then assemble to form edema toxin ( ET ) and lethal toxin ( LT ) , respectively [3 , 4] . ET and LT can enter a wide array of mammalian cells expressing either of two related surface receptors , CMG2 or TEM8 , where upon the toxins are internalized , leading to the release of the enzymatic A-subunits into the cytoplasm [5] . LF is a zinc metalloprotease that cleaves and inactivates mitogen-activated protein kinase kinases ( MAPKKs or MEKs ) to block MAPK signaling pathways [6] and , in some hosts , also cleaves NLRP1 to activate the inflammasome [7] . EF is a calmodulin-dependent AC , estimated to be more than a hundred times more potent than its mammalian counterparts in raising intracellular cAMP concentrations [8] . During the early stages of anthrax infection , LT and ET inhibit the innate immune response , reducing cell viability , disrupting chemotaxis and phagocytosis and deregulating cytokine production by macrophages , dendritic cells , and lymphocytes . These combined toxic effects promote bacterial growth and dissemination throughout the host [9 , 10] . In late fulminant stages of the disease , increasing amounts of ET [11] are released into the bloodstream , and in combination with LT cause edema , bleeding and hemorrhagic lesions ( ET ) , and atypical collapse of the cardiovascular system ( LT ) , often culminating in cardiac arrest and death [12 , 13] . Molecular pathways altered by the concerted effects of EF and LF were analyzed in transgenic Drosophila models by tissue-specific and conditional expression of the A-toxin subunit using the GAL4/UAS system [14] . Expression in the developing wing revealed that EF caused a phenotype very similar to that of a dominant-negative form of Rab11 , a small GTPase of the Rab subfamily essential for endocytic recycling [15 , 16] . Consistent with EF blocking Rab11-dependent trafficking , two known cargo proteins , Delta ( a transmembrane ligand activating the Notch receptor ) and the homophylic adhesion protein E-cadherin[17 , 18] failed to reach their normal destination at apical adherens junctions ( AJs ) . In addition , Rab11 levels were severely reduced in response to EF expression in the wing imaginal disc . This newly recognized activity of EF was also observed in mammalian cells , where ET caused a clear disruption of AJs and Notch signaling in several endothelial cell lines , and was essential for B . a . -induced vascular effusion in vivo [19] . To promote cargo vesicle fusion with the plasma membrane at proper apical sites , Rab11 relies on its effector Sec15 , which physically binds to the GTP-bound/active form of Rab11[13 , 20 , 21] . Sec15 is a key component of the exocyst , an octameric protein complex that triggers docking and SNARE-mediated fusion of cargo vesicles with the plasma membrane [22] . When over-expressed in various cell types , Sec15 promotes the assembly of large punctate structures[20] that also contain Rab11 , Sec15 , and other exocyst components . Consistent with previous observations , we found that EF prevented the formation of such Sec15-rich punctae . Interestingly , LF led to a similar inhibition of Sec15 punctae assembly , although via a Rab11-independent mechanism , indicating that Sec15 acts as a convergence point that integrates the effects of both anthrax toxins to block exocyst-mediated trafficking and disrupt integrity of the endothelial barrier [19] . Subsequent studies revealed that cholera toxin also blocks Rab11-mediated trafficking , an activity expected to increase intestinal epithelial permeability , paracellular water loss and diarrhea [23] . These similar cellular effects of ET and Ctx are likely to contribute to the hallmark pathological features and symptoms associated with anthrax and cholera respectively [24] . In the present studies , we delve deeper into the molecular pathways connecting ET-induced cAMP overload to inhibition of Rab11 . We apply a combination of approaches involving GTPase isoform-specific transgenes and antibodies , different Drosophila epithelial tissues , human cell lines , and in vivo experiments in mice . Our results indicate that EF disrupts Rab11-dependent processes after the GTP loading step . In flies , both cAMP effectors PKA and Epac disrupt Rab11-mediated junctional transport when artificially activated , but disable early versus late steps of the trafficking process , respectively . However , the Epac/Rap1 pathway seems to serve as the primary mediator of EF-induced toxemia in mammalian systems as well as in the Drosophila wing epithelium . Constitutive activation of Arf6 , a small GTPases involved in endocytic retrieval of junctional proteins [25] , causes phenotypes nearly identical to that of EF , and similarly alters Rab11 levels and distribution . These findings have potentially important practical implications , since chemical inhibition of Epac ( using the selective cAMP analog ESI-09[26] ) or Arf6 ( using SecinH3[27] or Slit[28] ) can reverse the effect of EF in a mouse footpad edema assay and in human cells . Such small molecule interventions open new potential therapeutic avenues for alleviating pathological effects of cAMP toxins and potentially other barrier disruptive agents .
In an effort better understand how EF blocks Rab11-dependent trafficking , we initially examined the behaviors of three YFP-tagged forms of Rab11: wild-type ( wt ) , activated ( * ) , and dominant-negative ( DN ) [29] . These variants were first expressed in the wing primordium in which inhibition of Rab11 by EF was initially discovered and analyzed [19 , 23] . The sub-cellular distribution of Rab11wtYFP detected by immuno-fluorescence appears as a grainy stain restricted primarily to the apical pole of epithelial cells ( Fig 1A ) . In addition to this wt pattern , activated Rab11 ( Rab11*YFP ) , a mutant that cannot hydrolyze GTP to GDP , displayed an additional staining component that accumulates at or near apical adherens junctions ( AJs ) ( Fig 1B ) . This latter staining is in line with the known role of Rab11 in junctional delivery ( see S1 Fig for co-localization of Rab11* and Drosophila E-cadherin , D-Ecad ) . We conclude that active GTP-bound Rab11 is selectively directed to cell-cell contacts at AJs . Consistent with this hypothesis , a dominant-negative Rab11 ( Rab11DNYFP ) , locked in its inactive GDP-bound conformation , did not display a preferential junctional distribution nor apical accumulation ( Fig 1C ) . We then turned our analysis to larval salivary glands , which are comprised of large polyploid secretory cells[30] , where the junctional-specific distribution of Rab11*YFP appears more pronounced ( Fig 1E ) . In these cells , over-expressed Rab11wtYFP was distributed throughout the cytoplasm , albeit excluded from densely packed secretory granules , with higher levels detected in the vicinity of intercellular junctions ( Fig 1D ) . Rab11*YFP behaved similarly but , in addition , exhibited a strong junctional staining component ( Fig 1E and 1H ) . In contrast , Rab11DNYFP did not concentrate at junctions , but altered the size and shape of secretory granules ( Fig 1G , thin arrows ) , suggesting that Rab11 normally plays a role in the formation or trafficking of these granules . These findings in the salivary gland confirm our results in wing discs suggesting that the activated GTP-bound form of Rab11 is selectively directed to AJs . According to the hypothesis that only the activated form of Rab11 traffics to junctions , factors blocking Rab11 upstream of the GTP-loading step should have no effect on Rab11* distribution , whereas inhibitory factors acting downstream of Rab11 should prevent Rab11* from accumulating at AJs . To test this model , we employed two RNAi constructs , one for knocking-down expression of Crag , which is the only known GEF specifically dedicated to activating Rab11 [31] , and the other for knocking-down Sec15 , an important Rab11 effector required for junctional delivery [18] . Specific inhibitory activities of these RNAi lines were confirmed using epitope-tagged forms of Crag and Sec15 ( See S2 Fig ) . When co-expressed with Rab11*YFP , Sec15-RNAi clearly prevented Rab11*YFP from reaching the AJs ( S3 Fig ) , consistent with Sec15 acting downstream of Rab11 activation . In contrast , Crag-RNAi had no effect on Rab11*YFP distribution , consistent with Crag acting upstream of Rab11 ( S3 Fig ) . Next , we examined whether EF blocks activation ( GTP loading ) of Rab11 or a subsequent step , by testing the effect of EF on Rab11*YFP localization . Expression of EF blocked all Rab11* junctional accumulation ( Fig 1F , compare panels 1H and 1I showing higher magnifications , see S4 Fig for quantifications of junctional Rab11* in response to EF expression ) . As the constitutively activated mutant Rab11*YFP remains sensitive to EF , we conclude that this toxin acts after the GTP-loading step . We next examined the behavior of endogenous Rab11 in salivary glands and its response to EF challenge using an antibody that detects all forms of Rab11 ( α-Rab11 ) . In wt glands , Rab11 shows a granular distribution with a higher concentration in the vicinity of cell junctions ( Fig 1J and 1L ) , which may represent an enrichment in activated Rab11 . In EF-expressing glands , this juxta-junctional staining was clearly reduced: Rab11 dots were detected at similar levels as in wt glands , but very few accumulated around the junctions ( Fig 1K and 1M ) . These findings are consistent with the hypothesis that EF prevents activated Rab11 from reaching the AJs . To test this model further we employed an antibody that specifically detects the activated Rab11 pool ( α-Rab11* ) . Consistent with the observations described above , we found that endogenous activated Rab11 localized predominantly to AJs ( Fig 1N and 1P ) . In EF-expressing glands , the overall levels of activated Rab11* were not obviously altered , however , less activated Rab11 accumulated at the AJs ( Fig 1O and 1Q , compare with 1N and 1P ) . Similarly , in EF-expressing discs , activated Rab11 levels remained comparable to wt levels , while junctional accumulation was severely reduced by EF ( Fig 1R and 1S ) . We conclude that EF does not interfere with Rab11 activation ( GTP loading ) , but instead blocks Rab11 function at a subsequent step ( s ) to prevent the activated form of Rab11 from trafficking to AJs . Next , we tested whether the association between Rab11 and its known cargo protein D-Ecad was affected by EF . As expected , co-labeling of Rab11 and D-Ecad in wt salivary glands revealed strong co-localization at cell junctions ( Fig 1T and 1U -wide arrow- ) and in punctate structures near the junctions ( Fig 1U -thin arrow- ) . In glands expressing EF , however , D-Ecad approached the cell surface ( Fig 1V and 1W ) , but failed to fully localize to AJs , as revealed by gaps in staining between cells ( Fig 1W , arrows ) . Similarly , expression of Rab11DN led to an accumulation of D-Ecad just under the junctions , while many gaps were visible between cells ( S5 Fig ) . These observations suggest that in salivary glands , Rab11 is not required for trafficking D-Ecad to the proximity of junctions , but is critical for the final delivery at the plasma membrane through vesicular fusion . Importantly , in EF-expressing glands , Rab11-DEcad co-localization was abrogated ( Fig 1V and 1W , wide arrows ) . We conclude that EF blocks the association between Rab11 and trafficking vesicles containing cargo proteins such as D-Ecad , leading to a failure in final step of junctional delivery with the consequence of weakened AJs . cAMP stimulates two main effectors: PKA and Epac , a GEF that activates the small GTPase Rap1 [32] [33] . We activated each branch of the cAMP pathway separately , using either a constitutively active form of PKA ( PKA* , consisting of the catalytic domain only [34] ) , or an activated form of Rap1 ( Rap1* , which is locked in its GTP-bound form [35] ) . We previously reported that both PKA* and Rap1* expressed in the wing primordium caused a reduction in Rab11 levels , blocked apical accumulation of Delta , and prevented the formation of Sec15 structures in the wing primordium [23] , suggesting that over-stimulation of each branch of the cAMP pathway can inhibit Rab11 . We thought to resolve the respective activities of PKA and Rap1 on Rab11 function further , by co-expressing the activated form of Rab11 ( Rab11*YFP ) with either PKA* or Rap1* . PKA* profoundly altered Rab11* distribution , both by eliminating accumulation of Rab11* at AJs ( Fig 2B and 2E , compare with 2A and 2D ) in a similar , albeit stronger , fashion to EF ( Fig 1F and 1I ) , and also by preventing the formation of secretory granules ( or dramatically reducing their size ) . These combined effects of PKA* result in Rab11* being ubiquitously distributed throughout the cytoplasm ( Fig 2B and 2E ) . A similar pattern was observed when staining for total endogenous Rab11 , which lost its tendency to concentrate around the junctions in response to PKA* expression ( Fig 2H , compare with 2G ) . Surprisingly , PKA* induced a strong increase in overall Rab11 levels in salivary glands , which is opposite to its effect in wing imaginal discs[19 , 23] . Consistent with Rab11-dependent trafficking being disrupted by PKA* , adherens junctions appeared weakened in PKA*-expressing glands , with more D-Ecad accumulating in the cytoplasm and around the AJs ( Fig 2K ) than in the wt glands ( Fig 2J ) . In contrast to PKA* , Rap1* expression in salivary glands did not prevent Rab11*YFP from accumulating near cell boundaries ( Fig 2C and 2F ) . However , instead of the typical single sharp line coinciding with cell junctions observed with Rab11*YFP alone ( Fig 2D ) , co-expression with Rap1* resulted in a double row of Rab11* staining , revealing a narrow gap between adjacent cells ( Fig 2F , arrows ) . This phenotype suggests a failure of the final fusion event between cargo vesicles and the plasma membrane . Consistent with these observations , endogenous total Rab11 staining was also concentrated in a sub-junctional zone in response to Rap1* expression ( Fig 2I , arrows ) , revealing narrow intercellular gaps . These Rap1*-expressing glands also showed an accumulation of small D-Ecad-rich vesicles near inter-cellular boundaries , while normal AJs failed to form ( Fig 2L ) . These results confirm the view that over-activation of each branch of the cAMP pathway can block Rab11-dependent trafficking , but that PKA* does so at an early step when vesicle loading takes place , while Rap1* acts later during the final vesicle delivery process . In adult Drosophila wings , both PKA* and Rap1* cause phenotypes similar to that of EF ( compare Fig 2N and 2O with 2Q ) consisting of smaller wings with blisters and thicker veins . The PKA* phenotype , however , is predominantly restricted to the center of the wing ( Fig 2N ) , while Rap1* , like EF , affects the entire wing blade ( Fig 2O and 2Q ) . Consistent with PKA* and Rap1* intersecting a common pathway , we found that co-expression of Rap1* and PKA* led to a drastically enhanced synergistic phenotype ( Fig 2P ) . While these gain-of-function studies reveal that both PKA and Rap1 signaling can interfere with Rab11 trafficking when artificially stimulated , we also tested which cAMP pathway might be required to mediate the effects of EF . We selectively blocked the Epac/Rap1 branch by expressing different EpacRNAi transgenes , which did not produce any notable phenotype on their own ( S6 Fig ) . When combined with EF , however , EpacRNAi significantly reduced the EF phenotype ( Fig 2R , compare with 2Q , see S6 Fig for quantifications ) . In contrast , reducing the levels of PKA-C1 , the major PKA catalytic subunit in Drosophila ( by two heterozygous loss-of-function PKA-C1 alleles ) , had little if any effect on the EF phenotype ( S6 Fig ) . We conclude that the Epac/Rap1 pathway is the predominant mediator of EF in the wing epithelium . In order to direct cargo vesicles to the AJs and promote their fusion with the plasma membrane , Rab11 must interact with several known effectors , including Rab11-FIPs ( Rab11 Family-Interacting Proteins [36] ) and Sec15 , a component of the exocyst complex that is critical for its assembly [20] . Drosophila has a single ortholog of Rab11-FIP ( dRip11 [37] ) , as well as unique representatives of all core exocyst components [18] . We first tested the effect of EF on Rab11 effectors by expressing a full length GFP-tagged dRip11 UAS transgene [37] in the salivary glands . When expressed alone , this fusion protein was strongly concentrated at cell junctions ( Fig 3A ) . Co-expression of EF with dRip11 reduced , but did not eliminate junctional accumulation of dRip11 and also resulted in forked and irregular cell borders ( Fig 3B ) . Because Rip11 is a Rab11-binding protein , we also examined association between these two proteins , which we visualized by expressing Rab11*YFP ( detected with a rat anti-GFP antibody ) and staining for the endogenous dRip11 . This particular double stain revealed frequent co-localization of the two proteins in bright dots in the vicinity of intercellular junctions ( Fig 3C , arrows in lower panel ) . Co-expression of EF with Rab11*YFP severely reduced its co-localization with Rip11 ( Fig 3D and 3G ) , supporting the hypothesis that high levels of cAMP trigger a dissociation of Rab11 and Rip11 , or prevent their initial association . Similarly , co-expression of PKA* with Rab11* also largely eliminated co-localization of Rab11* and Rip11 ( Fig 3E and 3G ) . Interestingly , Rap1* also affected this association: Rab11*YFP and Rip11 proteins remained present in adjacent but non-overlapping vesicles ( Fig 3F , quantifications in 3G ) , suggesting that both Rap1* and PKA* have effects on the Rab11*-dRip11 interaction albeit through distinct mechanisms . In contrast to full-length dRip11 , a truncated dominant-negative form of dRip11 ( dRip11DN ) retaining only the C-terminal Rab11-binding domain[37] , did not accumulate at cell-junctions in salivary glands , consistent with its N-terminal cholesterol-binding domain being essential for associated cargo vesicles to traffic to AJs . Instead , dRip11DN was distributed in a reticulated pattern throughout the cytoplasm , although it did show higher juxta-junctional levels ( S7 Fig ) . Small cytoplasmic Rab11 staining punctae strongly co-localized with dRip11DN-GFP ( S7 Fig ) , consistent with dRip11DN retaining its Rab11-binding domain . Interestingly , when co-expressed with EF , this punctate co-localization was not reduced , but rather transformed into rings that encircled secretory granules ( S7 Fig ) . Thus , EF does not abrogate association between Rab11 and dRip11DN . Because deletion of the first 700 aa of Rip11 ( a region containing a verified PKA phosphorylation site in humans [38] and several such predicted sites in Drosophila ) results in an EF-resistant association between Rab11* and dRip11DN , it is possible that PKA phosphorylation may contribute to this dissociation . We next examined the relationship between Rip11 and Rab11 in mammalian Madin-Darby canine kidney ( MDCK ) cells , in which the role of Rab11 in cadherin trafficking has been well established [39] . Co-expression of human Rab11-DsRed and EGFP-Rip11 constructs in these cells revealed strong co-localization throughout the cytoplasm , and a tendency for both proteins to accumulate at cell margins ( Fig 3H ) . Upon treatment with ET , however , we observed a significant reduction in Rab11 and Rip11 co-localization , and a reduction in Rab11 localization at the plasma membrane ( Fig 3I , see S8 for Pearson’s coefficient quantifications ) . Mirroring our observations in Drosophila salivary glands , EGFP-Rip11 accumulation at cell boundaries was reduced by ET-treatment ( Fig 3I ) . Interaction between endogenous activated Rab11 and its effectors was also tested in human brain microvascular cells ( HBMECs ) transfected with a mammalian Sec15-GFP construct . High-level expression of Sec15-GFP led to formation of punctate fluorescent structures ( S9 Fig ) , the formation of which depends on Rab11 [19] . Consistent with Sec15 associating with the active form of Rab11 , we detected , using an anti-Rab11* antibody , a high degree of co-localization between Sec15-GFP fluorescence and Rab11* . In this context of Sec15 over-expression , we also visualized co-localization of Rab11* with endogenous Rip11 ( S9 Fig ) . When these cells were treated with ET , Sec15-GFP punctae were significantly reduced after 6 hours , and the remaining punctae no longer co-localized with Rab11* or Rip11 ( S9 Fig ) . Cumulatively , these experiments suggest that EF-induced dissociation of Rab11* from its effectors Rip11 and Sec15 is a well-conserved process across species . Junctional homeostasis is also established by a balance of Rab11-mediated delivery of junctional cargo and retrieval of proteins via endocytic processes . Arf6 , a small GTPase of the Arf subfamily ( ADP-ribosylation factors ) is involved in early steps of endocytosis from the plasma membrane , exocytosis , and endosomal recycling , and is predominantly localized to the plasma membrane and endosomes [25] . Arf6 activation contributes to sepsis by promoting vascular leakage through excessive internalization of VE-cadherins [40] and additionally interacts directly with exocyst components [41] . We tested whether Arf6 also exerted a role in mediating the phenotypes caused by cAMP-producing toxins in our system by expressing an activated form of this small GTPase ( Arf6* ) . Strikingly , Arf6* caused a wing phenotype nearly identical to that induced by EF ( Fig 4A and 4B ) or Rab11DN [19] , consisting of small narrowed wings with thickened veins and blisters . In contrast , the wild-type form of Arf6 ( Arf6wt ) when expressed alone did not cause any detectable phenotype ( Fig 4D ) . However , both activated and wild-type forms of Arf6 strongly enhanced the EF wing phenotype ( Fig 4C and 4E ) . Further analysis revealed that Arf6* reduced the levels and apical restriction of Rab11 in the wing discs ( Fig 4H and 4I ) , diminished the formation of Sec15-rich structures , and reduced total Sec15 levels ( Fig 4J and 4K ) , in a manner similar to what we observed with EF [19] . Arf6* expression also reduced the levels of junctional and total D-Ecad ( Fig 4M , compare to 4L ) , as would be expected from its wing phenotype and effects on Rab11 and Sec15 . Given the striking similarities between Arf6* and EF phenotypes , we tested whether Arf6 contributes to mediating the effect of EF in the developing wing , making use of an Arf6-RNAi construct that is highly effective in suppressing Arf6 expression ( S2 Fig ) . Arf6-RNAi did not produce any noticeable phenotype on its own ( Fig 4F ) , but did exert a significant suppression of the EF wing phenotype ( Fig 4G , compare with 4A ) . Arf6-RNAi suppression of the EF phenotype was yet more pronounced at the level of junctional E-Cadherin expression ( Fig 4O , compare to 4N ) . We also tested whether Arf6* altered the distribution of Rab11*YFP in salivary glands . As observed with EF ( Fig 1F and 1I ) , Arf6* reduced the concentration of Rab11*YFP at the AJs ( Fig 4Q , compare to 4P ) , revealing that Arf6* similarly inhibits Rab11 at a step subsequent to GTP loading . In contrast to EF , however , Arf6* induced an intracellular accumulation of Rab11 , while also reducing Rab11 levels near the junctions ( Fig 4S , compare to 4R ) , and caused striking accumulations of D-Ecad below the apical plasma membrane ( Fig 4U , compare to 4T ) . In aggregate , these observations suggest that the Arf6 pathway inhibits Rab11 activity , but does so through a mechanism distinct from that of EF . As described above , activation of PKA* , Rap1* , and Arf6* mimic features of the EF phenotype in Drosophila . We wondered whether the same might be true in mammalian systems and thus examined the relative contributions of each of these pathways in EF-induced toxemia in various experimental models relevant to B . a . infection in mammals . In HBMECs , ET treatment reduced total cadherin levels and weakened AJs as indicated by staining with an anti-pan-cadherin ( p-Cad ) antibody ( Fig 5B , compare with 5A ) , as shown previously [19] . Western-blot analysis confirmed a drastic decrease in p-Cad and Rab11 levels in response to treatment with ET or dcAMP ( S11 Fig ) . Similar reductions in Rab11 levels in response to EF have been documented histochemically in Drosophila wing imaginal discs [19] . To inhibit the Arf6 pathway , we treated HBMECs with Slit2 , a secreted peptide that activates the Robo4 receptor to promote vascular stability via stimulation of the ArfGAP GIT[28] . Cells co-treated with ET and Slit2 appeared resistant to ET , as clearly illustrated by the robust rescue of junctional pan-cadherin accumulation ( Fig 5C ) . These findings suggest that Arf6 contributes to EF-induced inhibition of the Rab11/exocyst complex and weakening of AJs . Next , to determine the relative contribution of each branch of the cAMP pathway , we co-treated ET-intoxicated cells with ESI-09 , an inhibitor specific for Epac[26] , or with H89 , a well-characterized inhibitor of PKA[42] . We found that only ESI-09 could partially restore cadherin expression at AJs ( Fig 5D ) , although junctions did not appear as regular as in untreated cells . In contrast , H89 provided no obvious rescue to the ET-induced junctional phenotype ( Fig 5E ) . We conclude that Epac/Rap1 is the predominant pathway mediating the effects of ET on exocyst-dependent junctional cadherin trafficking in HBMECs . We next examined the relative contributions of the PKA and EPAC pathways as well as Arf6 in a quantitative in vivo footpad-swelling assay , in which intra-dermal injection of ET results in a robust and quantifiable edema ( Fig 5F ) [43] . In mice pretreated with SecinH3 , a compound that inhibits the Arf6-GEF ARNO thereby lowering Arf6 activity [28] , ET-induced swelling was strongly reduced ( Fig 5G ) . Indeed , in animals in which systemic pre-treatment with the drug induced observable symptoms of malaise ( presumably indicative of potent systemic pharmacological action ) , ET-induced edema was virtually abolished . We then examined contributions of the cAMP effector PKA and Epac to ET-induced edema , by comparing the relative abilities of H89 and ESI-09 to block ET-induced footpad swelling ( Fig 5F and 5H ) . Reinforcing the results of our experiments in flies and with HBMECs , we found that while ESI-09 virtually abolished ET-induced edema , H89 had little or no effect ( Fig 5H ) . We conclude that the Epac/Rap1 pathway is the primary mediator of EF-induced edema . In addition , we tested the effect of AG1024 [44] , an inhibitor of insulin-like growth factor receptor ( IGF-1R ) [45] . Because IGF-1R has been shown to indirectly stimulate Rap1 [46] , we hypothesized that inhibition of IGF-1R by AG1024 might conversely result in Rap1 inhibition . Indeed , pre-treatment of mice with AG1024 also led to significant reduction of edema , which was particularly strong at early time points ( Fig 5I ) . These findings , together with results described above provide a framework for how effector pathways contribute to cAMP-mediated disruption of Rab11-dependent membrane trafficking ( See Fig 6 for summary diagram ) . In addition to a prominent role of the Epac/Rap1 branch in mediating the effect of ET , our study reveals a previously unappreciated form of negative cross-regulation between the machineries responsible for the delivery versus retrieval of membrane bound cargo . Importantly , small molecule inhibitors such as SecinH3 , ESI-09 , and AG1024 offer potential for new therapeutic avenues for treating a range of diseases involving compromised barrier integrity of epithelial or endothelial sheets .
As is typical of small GTPases , Rab11 cycles between active ( GTP-bound ) and inactive ( GDP-bound ) conformations , the former permitting interaction with effector proteins to carry out downstream functions . Two types of regulators , activating GEFs and inactivating GAPs provide control for this essential cycle . In the particular case of Rab11 , Crag ( the Drosophila homolog of human DENND4A ) is the only known Rab11-dedicated GEF [31] . Similarly , only one Rab11-specific GAP has been identified: EVI5 [47–49] . Neither of these regulators contains an identified cAMP-binding domain that could provide a direct link between cAMP and upstream regulation of Rab11 . Consistent with this inference , we found that EF acted on Rab11 at a step subsequent to GTP loading . Indeed , transport of vesicles carrying the constitutively activated mutant Rab11*YFP were blocked by EF , while total endogenous levels of Rab11-GTP did not appear to be greatly altered . Association between Rab11 and its effectors Rip11 and Sec15 was abrogated by EF in several settings , including Drosophila salivary glands and human cells . The Rab11 effector Rip11 is an attractive candidate for mediating some of EF effects , as it contains a verified PKA phosphorylation site located in the central portion of the protein [38] . Indeed , PKA-dependent phosphorylation of Rip11 is required for cAMP-potentiated insulin secretion in pancreatic β-cells [38] . In addition , Ser/Thr phosphorylation is responsible for Rip11 transition from the insoluble to cytosolic fraction in intestinal CACO-2 cells [50] . Although it was not determined whether the latter modification was specifically PKA-dependent , this study proposed a model in which phosphorylation of Rip11 is essential for cycling to a free state following interaction with Rab11 and specific membrane compartments prior to its re-associating with Rab11 . Our data show that the association between Rab11 and Rip11 can be disrupted by EF in Drosophila and mammalian endothelial or embryonic kidney cells . It is possible that unrelenting phosphorylation of Rip11 by PKA may cause the premature dissociation of Rab11 and its effectors , potentially leading to a failure to reach the AJs . While this PKA-dependent phosphorylation of Rip11 has been demonstrated in human pancreatic cells , it is not known whether it occurs in Drosophila . As dRip11 contains 19 candidate PKA phosphorylation sites , further investigation will be necessary to determine whether phosphorylation of one or more of these sites occurs and promotes the dissociation between dRip11 and Rab11 . Intriguingly , Drosophila Sec15 also harbors several putative PKA phosphorylation sites , although such predicted sites are missing in its human counterpart . Importantly , we found that artificial stimulation of Rap1 also causes a loss in Rab11*/Rip11 co-localization resulting in correlated but separated staining foci of these two proteins , suggesting that the later acting Epac/Rap1 pathway may feedback on this process ( see below ) . The second branch of the cAMP pathway mediated by the cAMP-regulated GEF Epac and its partner Rap1 [32] contributes significantly to the effect of EF in flies , and surprisingly appears to play the predominant role in the mammalian systems we examined . In flies , activated Rap1 ( Rap1* ) causes a wing phenotype more similar to that of EF and Rab11DN than that of PKA* . We previously reported that Rap1* reduces the levels of Rab11 and prevents formation of Sec15 punctae [23] . In the present study , we find that blocking expression of Epac significantly reduces the intensity of the EF phenotype . In addition , Rap1* alters the distribution of Rab11* and inhibits Rab11*/Rip11 co-localization . We hypothesize that the final exocyst- and SNARE-dependent fusion event with the apical plasma membrane is subjected to inhibition by exuberant Rap1* activity , leading to accumulation of non-functional Rab11* just beneath the plasma membrane . Consistent with this hypothesis , Rap1 has been implicated by many studies in regulating of both cadherin and integrin-mediated cell-cell adhesion ( reviewed in [51] [52] [53] ) . Further indicating a functional connection between Rap1 signaling and Rab11-dependent trafficking , Rap1 and Rab11 over-expressed in human cells co-localize in a recent study [54] . Additional experiments will be required to elucidate the molecular interactions connecting the activities of these two GTPases . The small GTPase RalA is a possible candidate for mediating the activity of Rap1 , through activation of the Rap1 effector Rgl1 , a positive regulator ( GEF ) of RalA . Because lowering the dose of Rgl1 , or expressing a dominant-negative form of RalA , can suppress Rap1*-induced phenotypes in Drosophila , it has been proposed that RalA may act downstream of Rap1 [35] . Also , RalA is known to directly bind to exocyst components Sec5 [55] [56] [57] and Exo84 [58] and plays a central role in regulating exocyst-mediated processes in several settings , including the release of Von-Willebrand Factor from endothelial cells , or insulin secretion in pancreatic β-cells ( reviewed in [51] and [59] ) . In addition , a recent study identified Arf6 as a key component acting downstream of RalA , mediating its effect on exocyst-dependent delivery of raft micro-domains to the plasma membrane [60] . Thus , RalA over-activationmay contribute to mediating the effect of cAMP toxins on exocyst inhibition downstream of Rap1 , although this hypothesis needs to be tested in future experiments . We previously showed that EF caused a drastic reduction in total Rab11 levels in wing epithelial cells[19] . Here , we find that this effect is also evident in HBMECs treated with ET , but is dependent on cell context , since inhibition of Rab11 function can be uncoupled from reduction in total Rab11 levels in Drosophila salivary glands . This reduction in Rab11 levels is unlikely to derive from transcriptional inhibition , as infection of HBMECs with B . a Sterne did not result in any change in levels of Rab11 transcripts ( Nina Van Sorge , personal communication ) . Similarly , in Drosophila wings , where EF also triggers great reduction in Rab11 protein levels , mRNA transcript levels again were not greatly affected ( Valentino Gantz , personal communication ) . In HBMECs , where Rab11 levels are reduced by ET treatment , we observed that total levels of cadherins were also severely reduced in ET-treated cells . Although the precise mechanism responsible for the loss of these proteins following ET treatment remains to be explored , it is worth noting that degradation of VE-cadherins has been observed following silencing of Rab11 in human endothelial cells [61] , in which Rab11 is important for stabilizing cadherins at the AJs . Thus , it is possible that following EF intoxication , Rab11 and cadherins are routed to the lysosomal pathway and degraded , further impairing endocytic recycling and junctional integrity . Such an attractive hypothesis could explain the catastrophic loss of cadherins observed in ET-treated cells . Numerous studies have demonstrated the positive role of physiological induction of cAMP in junction establishment and stabilization , through stimulation of both PKA and Epac [38 , 62] . It may therefore seem counterintuitive that cAMP produced by EF or other toxins may exert an opposing effect and jeopardize junctional integrity . In principle , high versus low concentrations , sustained versus transient production , and perinuclear vs cortical subcellular distribution of toxin-delivered cAMP could elicit such opposite outcomes . In the particular case of Rab11-dependent trafficking , low physiological levels of cAMP may exert their positive effects by promoting the release of Rip11 from Rab11 , as necessary to allow the final fusion event between recycling endosomes and the plasma membrane . In contrast , pathologically elevated cAMP concentrations may cause premature dissociation of the Rab11-Rip11 complex and permanently block that cycle . Similarly , uncontrolled stimulation of Rap1 by Epac could also have a negative impact on junctional transport: titration of critical partners , failure to return to complete the necessary GTP/GDP cycle , or negative feedback interference with other important steps , could explain the occurrence of this apparent paradox . Another molecule potentially at play during the response to cAMP is the small GTPase RhoA . RhoA can be phosphorylated by PKA , which inhibits its activation and prevents increased endothelial permeability during inflammation [63] , the potential interplay between RhoA and the exocyst downstream of cAMP signaling in EF-intoxicated cells also merits further examination . The small GTPase Arf6 initiates retrieval of membrane proteins from cell junctions in a wide variety of cells types [25] . Arf6 , a member of the ADP-ribosylation factor subfamily , is located at the plasma membrane and some endosomal compartments , and is involved in endocytosis from the plasma membrane , vesicular recycling , and exocytosis [64] . Importantly , Arf6 plays a role during sepsis to mediate acute VEGF-induced vascular permeability [40 , 65] . Whether linchpin regulators of opposing vesicular trafficking pathways such as Arf6 and Rab11 interact had not yet been extensively explored . In this study , we present evidence that these trafficking systems do in fact engage in cross-inhibitory interactions . Consistent with the published role of Arf6 in promoting VE-cadherin endocytosis [66] , the activated form of Arf6 ( Arf6* ) caused phenotypes similar to those of EF . Our findings suggest that the activity of Arf6 negatively feeds back on vesicular transport to the plasma membrane by inhibiting Rab11 function . Previous studies showed that Arf6 physically interacts with the exocyst component Sec10 [41] , defining a possible avenue for our observed effects of Arf6 on Rab11 levels and distribution . Given the negative regulation of Rab11 by Arf6 in flies and its known role in compromising barrier function in the mammalian vasculature during sepsis [28 , 40] , we tested whether inhibitors of this pathway might antagonize the effects of EF . In human endothelial cells , we indeed found that treatment with Slit2 , a secreted peptide indirectly blocking Arf6 function , could reverse the effects of EF , restoring junctional integrity . Similarly , pharmacological inhibition of Arf6 by SecinH3 , a compound that inhibits the ArfGEF ARNO , potently blocked EF-induced edema in a mouse footpad assay . An emerging lesson from the current and prior studies is that blocking multiple steps of branching pathways that converge on critical nodes in endocytic recycling may allow pathogens to weaken host protective mechanisms that rely on junctional integrity [24] . For example , LF , the other toxic factor secreted by B . a , blocked exocyst-mediated vesicular docking downstream of Rab11 via inhibition of MAPK signaling . It will be interesting to explore how the various effects of EF and LF are integrated to achieve an efficient inhibition of junctional delivery , and if any compound identified in this study can also block some of the downstream effects of LF . Altogether , our study suggests that a broad range of barrier disruptive diseases ranging from cAMP related toxemia to inflammatory autoimmune diseases that involve positive feedback loops between immune activation and barrier disruption , could potentially be treated with compounds that inhibit Arf6 or Epac/Rap1 , or by yet undiscovered compounds that may boost Rab11 activity .
All experiments were performed in strict accordance with guidelines from the National Institute of Health and the Animal Welfare Act , approved by the Animal Care and Use Committee of University of California , San Diego and the National Institute of Allergy and Infectious Diseases , National Institutes of Health ( approved protocols s00227m and LPD-8E ) . Anesthesia and euthanasia were performed using Isoflurane and CO2 , respectively . All efforts were made to minimize suffering of animals employed in this study . UAS-EF construct and line were described previously[19 , 23] . UAS-Rab11wtYFP/TM3 ( #9790 ) , UAS-Rab11*YFP ( 3rd chr . # 9791 ) , UAS-Rab11DNYFP ( #23261 ) , UAS-CragRNAi ( 2nd chr . #53261 ) , UAS-CragHA ( 3rd chr . #58463 ) , UAS-Arf6RNAi ( 3rd chr . #51417 ) , UAS-PkaDN/CyO ( #5282 ) , and Pka-C1B10 ( #32018 ) lines were obtained from Bloomington Drosophila Stock Center ( BDSC ) . UAS-EpacRNAiv50272 and UAS-EpacRNAiv50273 were obtained from Vienna Drosophila Resource Center ( VDRC ) . UAS-Rap1*/TM6 and UAS-PKA*/CyO were generated by I . Hariharan ( UCB ) , and D . Kalderon ( Columbia University ) , respectively . UAS-Rip11GFP and UAS-Rip11DNGFP were kindly provided by Don Ready ( Purdue University ) . UAS-Arf6* was generated in the Olson laboratory ( UT Southwestern ) . Imaginal discs were dissected , fixed and stained using standard procedures . Salivary glands were dissected similarly , fixed for 30 minutes , and left attached to carcasses until ready to mount in SlowFade ( LifeTechnologies #S36936 ) , using double sided tape as a spacer to prevent tissue squashing . Antibodies: rabbit anti-GFP antibody ( 1/500 , ThermoFisher #A6455 ) , rat-anti GFP antibody ( 1/500 , SCBT #sc-101536 ) , mouse anti-Rab11 ( 1/200 , BD Biosciences #610657 ) , mouse anti Rab11-GTP ( 1/100 , NewEast Biosciences #26919 ) , D-Ecad ( 1/500 , DSHB #DCAD2 ) . The rabbit anti-Rip11 ( 1/1000 ) was a gift from D . Ready ( Purdue University ) and A . Satoh ( Hiroshima University , Japan ) , and guinea pig anti-Sec15 ( 1/1000 ) was kindly provided by Hugo Bellen ( Baylor College of Medicine ) . Images were collected by confocal microscopy on a Leica TCS SP5 . All images were acquired using a 40X or 63X objective , and all higher magnifications were obtained using a 4X digital zoom . Co-localization quantifications in Fig 3 used the coloc2 tool in ImageJ . MDCK cells ( ATCC CCL-34 ) were maintained in DMEM ( Corning; Manassas , VA ) containing 10% FBS , 1% Penicillin/streptomycin , 2 mM L-glutamine and were incubated in 37°C , 5% CO2 atmosphere . Cells were gently dislodged with 0 . 05% trypsin ( Mediatech ) and were electroporated with cDNA expressing DsRed-Rab11A ( Addgene ) and Rip11-EGFP ( Kind gift from Dr . Rytis Prekeris , Univ . of Colorado , Denver ) using Neon Transfection system ( Life Technologies ) according to manufacturer’s protocol . Briefly , cells were rinsed once with PBS and resuspended at a density of 107 cells/ml . cDNA expressing DsRed-Rab11A and Rip11-EGFP were added to the suspension , and cells were electroporated with a 10 μl Neon tip at 1650 V , 20 ms width and 1 pulse . Cells were transferred to 600 μl pre-warmed medium of which 300 μl cell suspension was plated on each well of 8 chamber tissue culture treated glass slide ( BD Falcon , Bedford , MA ) . Cells were treated with 10 μg/ml EF +20 μg/ml PA for 4 h before fixation with 4% para-formaldehyde in PBS for 30 min at 37°C and processed for imaging . Fluorescence images were collected using a Delta Vision RT microscope . Colocalization between Rab11 and Rip11 was determined by measuring the Pearson's correlation coefficient ( PCC ) using the Velocity 6 . 3 imaging and analysis software ( PerkinElmer ) . Costes automatic thresholding method [67] was applied for background discrimination . HBMEC cultures were maintained in DMEM ( Corning ) containing 10% FBS , 1% Penicillin/streptomycin , 2 mM L-glutamine , and were incubated in 37°C , 5% CO2 atmosphere . Cells were gently dislodged with 0 . 05% trypsin ( Mediatech Inc . ) and cultured on glass poly-D-lysine coated chamber slides ( BD Falcon #354108 ) . At about 80% confluence , EF and PA ( 0 . 2 μg/ml and 0 . 4μg/ml , respectively ) were added to cells . Drug co-treatments included: Slit2 ( 10μg/ml , R&D systems # 8616 ) , ESI-09 ( TOCRIS #4773 , 100μM ) , and H89 ( TOCRIS #2910 , 10μM ) . After 24 h ( Fig 5 ) , cells were fixed for 10 min at -20°C in 100% Methanol , then washed with 0 . 1% Triton in PBS . Cells were stained with a mouse anti pan-Cadherin antibody ( Abcam , clone CH-19 , 1/100 ) . For S6 Fig , transfection of the Sec15-GFP was performed with the FuGENE 9 transfection reagent ( Roche ) according to manufacturer recommendations . Cells were treated with ET ( 2 μg/ml EF and 4 μg/ml PA ) , and fixed after 6hrs of treatment for 30 mins in 4% paraformaldehyde in PBS . Cells were stained with rabbit anti Rip11 ( Novusbio #NBP1-81855 , 1/500 ) and mouse anti Rab11-GTP ( NewEast Bioscience #26919 , 1/100 ) antibodies overnight at 4°C . Coverslips were washed , and incubated with secondary antibodies before mounting with Prolong Gold with DAPI mounting media ( ThermoFisher ) . BALB/cJ mice ( 8–10 weeks old , female; Jackson Laboratories ) were intraperitoneally injected with drugs . SecinH3 ( TOCRIS #2849 ) , AG1024 ( Selleckchem , S1234 ) , ESI-09 ( TOCRIS #4773 ) or H89 ( TOCRIS #2910 ) , or with vehicle ( 70% DMSO in isotonic glucose for SecinH3 , 30% DMSO in isotonic glucose for other drugs ) 2–3 h prior to injection of ET ( 0 . 15 μg/20 μl , right footpad ) or PBS ( 20 μl , left footpad ) , and in the case of SecinH3 , also 2 h post toxin injection . ESI-09 , AG1024 and H89 were administered at 10 mg/kg and SecinH3 was administered as 250 μl of 2 . 5 mM solution . Edema was assessed at 8–10 h , and 18–24 h by dorsal/plantar measurements using digital calipers . Untreated or ET intoxicated ( 24h ) HBMEC cells were lysed in RIPA buffer ( Cell Signaling Technology ) supplemented with mammalian protease inhibitor cocktail ( Sigma Aldrich ) . The lysates were clarified by centrifugation at 1000 g for 10 min at 4°C and LDS sample buffer ( NuPAGE ) was added . Samples were boiled at 95°C , run on a 4–12% SDS polyacrylamide gel ( Life Technologies ) and transferred onto PVDF membrane ( Bio-RAD ) . After incubation with primary antibodies against Rab11 ( #71–5300 , Thermo Scientific ) , Cadherins ( CH-19 , Abcam #ab6528 ) , and actin ( sc69879 , Santa Cruz Biotechnology ) , blots were probed with respective HRP conjugated secondary antibodies and developed using SuperSignal West Pico chemiluminescent substrate ( Thermo Scientific ) .
|
Recent anthrax outbreaks in Zambia and northern Russia and biodefense preparedness highlight the need for new therapies to counteract fatal late-stage pathologies in patients infected with Bacillus anthracis . Indeed , two toxins secreted by this pathogen—edema toxin ( ET ) and lethal toxin ( LT ) —can cause death in face of effective antibiotic treatment . ET , a potent adenylate cyclase , severely impacts host cells and tissues through an overproduction of the ubiquitous second messenger cAMP . Previously , we identified Rab11 as a key host factor inhibited by ET . Blockade of Rab11-dependent endocytic recycling resulted in the disruption of intercellular junctions , likely contributing to life threatening vascular effusion observed in anthrax patients . Here we present a multi-system analysis of the mechanism by which EF inhibits Rab11 and exocyst-dependent trafficking . Epistasis experiments in Drosophila reveal that over-activation of the cAMP effectors PKA and Epac/Rap1 interferes with Rab11-mediated trafficking at two distinct steps . We further describe conserved roles of Epac and the small GTPase Arf6 in ET-mediated disruption of vesicular trafficking and show how chemical inhibition of either pathway greatly alleviates ET-induced edema . Thus , our study defines Epac and Arf6 as promising drug targets for the treatment of infectious diseases and other pathologies involving cAMP overload or related barrier disruption .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"toxins",
"vesicles",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"enzymology",
"animals",
"toxicology",
"toxic",
"agents",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"signs",
"and",
"symptoms",
"experimental",
"organism",
"systems",
"edema",
"cellular",
"structures",
"and",
"organelles",
"drosophila",
"digestive",
"system",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"proteins",
"exocrine",
"glands",
"guanosine",
"triphosphatase",
"insects",
"cell",
"membranes",
"arthropoda",
"biochemistry",
"cell",
"staining",
"eukaryota",
"diagnostic",
"medicine",
"hydrolases",
"anatomy",
"cell",
"biology",
"salivary",
"glands",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2017
|
Anthrax edema toxin disrupts distinct steps in Rab11-dependent junctional transport
|
In a broad variety of bilaterian species the trunk central nervous system ( CNS ) derives from three primary rows of neuroblasts . The fates of these neural progenitor cells are determined in part by three conserved transcription factors: vnd/nkx2 . 2 , ind/gsh and msh/msx in Drosophila melanogaster/vertebrates , which are expressed in corresponding non-overlapping patterns along the dorsal-ventral axis . While this conserved suite of “neural identity” gene expression strongly suggests a common ancestral origin for the patterning systems , it is unclear whether the original regulatory mechanisms establishing these patterns have been similarly conserved during evolution . In Drosophila , genetic evidence suggests that Bone Morphogenetic Proteins ( BMPs ) act in a dosage-dependent fashion to repress expression of neural identity genes . BMPs also play a dose-dependent role in patterning the dorsal and lateral regions of the vertebrate CNS , however , the mechanism by which they achieve such patterning has not yet been clearly established . In this report , we examine the mechanisms by which BMPs act on cis-regulatory modules ( CRMs ) that control localized expression of the Drosophila msh and zebrafish ( Danio rerio ) msxB in the dorsal central nervous system ( CNS ) . Our analysis suggests that BMPs act differently in these organisms to regulate similar patterns of gene expression in the neuroectoderm: repressing msh expression in Drosophila , while activating msxB expression in the zebrafish . These findings suggest that the mechanisms by which the BMP gradient patterns the dorsal neuroectoderm have reversed since the divergence of these two ancient lineages .
In both Drosophila melanogaster and vertebrates , Bone Morphogenetic Proteins ( BMPs ) are expressed in the epidermal ectoderm abutting the dorsal border of the neuroectoderm [1] . The genetic network that underlies formation of a centralized nervous system consisting of segregated motor and sensory centers appears to have been conserved across bilaterians ( animals with right-left symmetry ) [2] . BMPs are thought to exert a common function in the early epidermal ectoderm during neural induction ( i . e . , suppressing expression of neural genes in epidermal regions that experience peak BMP levels ) . BMP signaling also acts subsequently in a dose dependent fashion to pattern dorsal versus medial regions of the neuroectoderm . For example , the trunk Central Nervous System ( CNS ) of both invertebrates and vertebrates consists of three primary rows of neuroblasts that are determined by the expression of three conserved transcription factors . In metazoan species spanning all three primary branches ( e . g . , Ecdysozoa -Drosophila , lophotrochozoa – annelids , and deuterostomes - vertebrates ) “neural identity” genes ( vnd/nkx2 . 2 , ind/gsh and msh/msx ) are expressed in the same relative order and orientation with respect to the dorsal-ventral axis and an epidermal BMP source . Moreover , in a broad range of organisms , BMPs and opposing antagonists have been found to play a key role in patterning the ectoderm and establishing neuronal fates . These commonalities suggest an ancestral origin for the CNS among bilateria [1]–[4] and raise the possibility that BMPs play a conserved role in patterning the CNS axis . Despite their consistent role in promoting epidermal over neuronal cells fates in diverse species , BMPs and other extracellular factors are deployed in diverse patterns and may act by distinct mechanisms to achieve D/V patterning [5] . For instance , in Drosophila , BMPs originating in the presumptive epidermis act to repress expression of neural genes during both neural induction [6] and subsequent neuroectodermal patterning [3] , [7] . In vertebrates , however , the prevailing view is that BMPs act as they do in flies to repress expression of neural genes within epidermal regions early during neural induction [1] , [8] but switch function later to activate expression of orthologous neural identity genes in dorsal regions of the neural tube ( e . g . , the msh orthologs Msx1/2 ) [9] . Thus , in mice , ectopic BMP signaling leads to ventral expansion of msx expression in the neural tube [10] . In contrast , in Drosophila , the absence of BMPs leads to msh expanding dorsally into non-neural domains [11] . In zebrafish , there is evidence that BMPs act in a bimodal fashion where intermediate BMP levels are necessary for activating Msx genes , while both low and high levels of BMPs repress or fail to activate these target genes [12] . Similarly , in amphioxus , a basal chordate , msx is expressed more broadly but at reduced levels in response to ectopic BMP signaling [13] . In Echinoderms , where BMPs and chordin are co-expressed in the ventral ectoderm that gives rise to neural tissue [14] , msx is expressed dorsally and is activated by peak levels of BMPs that diffuse dorsally from their ventral source into non-neural regions while Chordin remains restricted to ventral regions where it blocks the BMP response in neural cells [15] . While these conserved suites of gene expression strongly suggest a common ancestral origin for BMPs in axial patterning , it is unclear whether the regulatory mechanisms establishing these patterns have been similarly conserved during evolution . BMPs signal via hetero-tetrameric receptor complexes consisting of two type-I and two type-II subunits , which in turn phosphorylate the cytoplasmic transducing-SMAD proteins ( Mothers Against Dpp ( Mad ) in Drosophila , SMAD1/5/8 in vertebrates ) . Once phosphorylated , pMad/pSMAD1/5/8 translocates into the nucleus in a complex with Medea/Smad4 whereupon they act as transcription factors to regulate expression of BMP target genes ( reviewed in [16] ) . Mad and Medea ( Med ) bind DNA as a heteromeric complex consisting of two Mad subunits and one Med subunit to regulate genes through interactions with binding sites composed by a Mad ( GC-rich ) site separated , by a variable length spacer , from a Med ( Smad Binding Element or SBE ) site . One of the best characterized such sites in Drosophila is the brinker ( brk ) Silencer Element ( SE ) which has a spacer length of 5 nucleotides [17]–[20] . Brk encodes a transcriptional repressor protein and the brk gene itself is repressed by Dpp ( the Drosophila BMP4 homologue ) signaling . Repression of brk through its SEs requires the presence of the zinc-finger protein Schnurri ( Shn ) [21]–[23] , which is provided maternally and is also expressed zygotically in dorsal epidermal regions of the early embryo . Hence , in Drosophila , genes that are repressed by BMPs have been found to have binding sites for pMad/Med/Shn ( henceforth , pMMS ) complexes in their cis-Regulatory Modules ( CRMs ) while genes that are directly activated by BMPs , such as the inhibitory SMAD daughters-against-dpp ( dad ) , contain activating elements ( AE ) in their CRMs [24] . These AE elements also share a bipartite configuration ( GC-rich/spacer/SBE ) , but have configurations ( spacing and sequence constraints ) that do not allow for Shn binding and lead instead to the recruitment of activating transcriptional co-factors . Here , we compare BMP-mediated regulation of CRMs controlling the expression of the Drosophila msh and zebrafish and mouse msx genes in the early dorsal nerve chord . We identify zebrafish and mouse msx neuroectodermal CRMs that drive expression in the dorsal neuroectoderm . We find that both Drosophila msh and zebrafish msxB CRM-reporter transgenes respond to BMPs and characterize BMP responsive sites within these elements . Consistent with prior genetic studies [7] , the Drosophila msh CRM contains Shn-dependent SE sites that are required for BMP repression . Surprisingly , it also harbors sites that resemble known BMP-responsive activation sites , which , however , do not bind to pMad/Medea ( pMM ) complexes in vitro , but are nonetheless required for msh expression . In addition , we characterize a single SMAD binding site with a novel spacing of SMAD1/5/8 and SMAD4 binding motifs in a minimal zebrafish msxB CRM that is required for dorsal neuroectodermal expression . This comparison suggests that while overall gene expression patterns have been conserved between flies and zebrafish and are both regulated by BMP signaling , distinct mechanisms have evolved to generate the shared output patterns in these two widely separated metazoan lineages .
A 700 bp msh CRM ( henceforth referred to as ME for Msh Element ) has been identified that is directly repressed by Ind [25] . The response of the ME to BMP-mediated regulation has not yet been investigated , however . As is the case for the endogenous msh gene ( Fig . 1A ) , the expression of a ME-lacZ construct expands throughout the dorsal region of the embryo in dpp- mutants ( Fig . 1B ) . In order to determine whether Dpp regulates msh directly or indirectly , we analyzed BMP regulation of the ME element . Consistent with a direct role of BMP signaling on this CRM , genome wide chromatin immune precipitation ( ChIP ) data [26] , [27] revealed DNA binding sites for the BMP effectors Mad , Medea and Shn within the ME region in blastoderm stage embryos ( available on the UCSC genome browser - http://genome . ucsc . edu/ or the Berkeley Drosophila Transcription network Project - http://bdtnp . lbl . gov/Fly-Net/chipchip . jsp ) ( Fig . 1D ) . We confirmed the involvement of Shn in regulating msh within the neuroectoderm by examining homozygous zygotic shn- mutant embryos , which exhibit a partial dorsal expansion of msh expression ( Fig . 1C ) . To identify BMP responsive sites within the ME , we first scanned this element for known consensus binding sites for Mad , Med , Shn , and Brk . The two best characterized BMP responsive elements are the Silencer Element ( SE ) , which binds a trimeric complex comprised of pMMS ( GNCGNC ( N ) 5GNCTG ) , and the activator element ( AE ) , which binds pMM heteromers ( GGCGCCA ( N ) 4GNCV ) . Brk binding sites ( ( T ) GGCGYY ) overlap with a subset of AE elements [24] . Although there are no perfect consensus SE , AE , or Brk sites within the ME , we identified several candidate sites with either single base-pair mismatches to the SE or AE elements or variable spacer length ( N ) 5–6 . We defined three such candidate SE sites ( SE1 , SE2 and SE3 ) with a single nucleotide mismatch and two conserved candidate AE sites with a spacer of 6 nucleotides ( conforming to the expanded consensus: GNCGNC ( N ) 6GNCV ) and tested each of these sites for direct DNA binding of pMM or pMMS complexes in vitro using Electrophoretic Mobility Shift Assays ( EMSAs ) . The SE1 and SE2 candidate silencer sites ( Fig . 2A ) both conform to the relaxed consensus of GNYGNC ( N ) 5GNCTG ( where Y can be either C or T ) . EMSA experiments using DNA oligonucleotide probes reveal that pMM and pMMS complexes assembled on the SE1 and SE2 sites in a BMP dependent fashion ( Fig . 2B ) but not on the SE3 site ( Fig . S1C ) . As expected , mutation of the Med ( SBE ) motif within the SE1 ( SE1SBE ) or SE2 ( SE2SBE ) sites abolished binding of all BMP responsive complexes in vitro . In contrast , none of the candidate AE or Brk sites bound pMM , pMMS , or Brk complexes ( Fig . S1C ) ( see below however , regarding effects of mutating or deleting the candidate AE sites ) . In order to test the in vivo roles of the SE sites , we mutated each site ( i . e . , using the same SBE mutations that abolished all BMP responsive DNA binding in vitro described above ) and generated a series of small deletions spanning virtually the entire ME ( i . e . , all but 36 bp ) . These mutant constructs were inserted into the same chromosomal integration site as the reference ME construct using the PhiC31 transgenesis system [28] . Deletion of the 5′ most 100 bp of ME , which contains both SE sites , led to dorsal expansion of reporter gene expression ( Fig . 2C ) . Transgene expression , however , was also weaker within its normal neuroectodermal domain , suggesting that contributing activation sites are also present within this region . Targeted mutation of the individual SE1 and SE2 sites also led to discernable dorsal expansion of reporter gene expression , which was more pronounced for the SE2 mutant . Mutating both SE sites in combination ( SE1 , SE2 double mutant ) resulted in more prominent dorsal expansion than observed for either mutant alone , but still less than that observed for the wild-type ME ( or the endogenous msh gene ) crossed into a dpp- mutant background . We conclude that SE elements mediate direct BMP-dependent repression of the ME and that additional direct or indirect BMP-dependent inputs also contribute to negatively regulating this CRM . Our prior genetic studies revealed that BMP signaling is more effective in repressing expression of ind than msh [7] . One possible explanation for this differential response is that the ind CRM might contain higher affinity SE sites than those in the msh CRM . Indeed , a single perfect consensus matching SE site in the ind CRM ( Fig . 2B ) has been shown to be required for repression of this element dorsally [20] , [29] . In line with the possibility that SE sites in the ind and msh CRMs have differing affinities for binding pMMS complexes , modifying the SE2 site by one base-pair to adhere to the optimal SE consensus resulted in greater pMMS binding ( Fig . 2B - SE2* ) , which was most evident in competition experiments ( Fig . S1D ) . We tested whether the optimized ind-like SE2* site would result in repression of msh CRM activity in vivo . In support of this site being more effective at recruiting repressive pMMS complexes , reporter gene expression driven by the SE2* ME was greatly reduced relative to that of the wild-type ME . This reduced expression was BMP-dependent since SE2*ME-driven reporter gene expression was restored and expanded throughout the dorsal region in a dpp- mutant background to a degree comparable to that observed for the intact ME ( Fig . 2C ) . Taken together , these results suggest that differential affinities of pMMS complexes for SE sites in the ind and msh CRMs contribute to the mechanism by which silencer elements mediate graded BMP responses of these two genes in the Drosophila neuroectoderm . As mentioned above , in our initial search for BMP-responsive sites in the ME we identified two sites that were similar to activation elements ( AE ) but that did not bind pMM complexes in EMSA assays ( Fig . S1C ) . We nonetheless tested for potential roles of these sites by deleting them or creating a point mutation in one of them ( AE2 ) . Deletions encompassing either AE1 or AE2 or the AE2 point mutation greatly reduced ME-lacZ expression ( Fig . 3B , C , E ) , while deletion of the 3′ most region containing a previously reported Ind site [25] resulted in ventral expansion of reporter gene expression as expected . We tested the possibility that activation of the ME via AE2 might be balanced against repression mediated by the SE1 and SE2 sites by constructing a triple mutant in which all three sites were eliminated . We reasoned that if the AE2 site , which acts as a bonafide activation site , functions in a BMP independent manner , combining it with the double SE site mutant might result in loss of expression ( the AE mutant phenotype ) . On the other hand , if the AE2 site were providing an important activation function in the neuroectoderm via BMP signaling , the triple mutant should at least show ectopic expression dorsally ( e . g . , if this was a BMP-dependent activation site , relieving repression would give rise to normalized expression since we would be removing both activating and repressing components ) . We found loss of expression in this triple mutant comparable to that of the AE2 single mutant ( Fig . 3F ) , suggesting that activation via the AE2 site is BMP-independent . Although the above analysis suggests that the AE2 site acts in a BMP-independent fashion , we further examined the possibility that BMPs might play an activating as well as repressive role in regulating msh expression . Embryos that are dorsal- ( maternal ) ; dpp- ( zygotic ) double mutants express msh ubiquitously [11] . To test whether there might be a threshold at which Dpp enhances rather than suppress msh expression , we attempted to augment msh expression locally by generating embryos that lack Dorsal and whose only source of Dpp is one copy of dpp driven in a narrow stripe by the eve 2 CRM ( Fig . 3G , H ) or by , adding progressive amounts of Dpp ( by varying copy number of the dpp locus – Fig . S2A , B ) . In both cases , we observed only a diminution in msh expression , further arguing against any activating role for Dpp . Finally , we considered the possibility that BMPs might act indirectly to regulate msh expression via non-canonical mechanisms ( e . g . via ETS or the HMG-box Cic transcription factors ) by altering EGF-R signaling . We found no evidence , however , for a role of EGFR signaling in influencing the position of the dorsal border of msh expression ( Fig . S2C-E ) . In aggregate , our experiments suggest that BMP-dependent regulation of the ME is mediated by SE sites and by additional inhibitory inputs , which may act either directly or indirectly . The above analysis of the msh CRM in Drosophila is consistent with genetic data indicating that BMPs act by dosage sensitive repression of neural identity gene expression [7] . To determine the mechanism by which BMPs regulate expression of orthologous vertebrate Msx genes we sought to identify the zebrafish ( Danio rerio ) msxB CRM using the powerful tol2 transgenesis system [30] . We choose to focus on regulation of the msxB gene among the zebrafish Msx paralogs as this gene has the earliest onset and most specific pattern of expression in the dorsal neuroectoderm [31] . We identified a 2 . 4 Kb region of DNA immediately upstream of the zebrafish msxB coding region that drives faithful reporter gene expression in the dorsal neuroectoderm in both neural plate and early neural tube stages ( i . e . , 3–6 somite stage embryos ) of a stable transformant line ( Fig . 4A , B ) . This fragment has two peaks of strong sequence conservation among vertebrates , which overlap regions of predicted open chromatin [32] ( Fig . 4A ) . Later during neural tube stages , the early neural plate expression pattern fuses into a single dorsal zone ( e . g . , top panels in Fig . 4D ) . We also tested a 5 Kb genomic fragment upstream of the mouse ( Mus musculus ) Msx1 gene , which like the zebrafish msxB CRM carries sequences lying immediately upstream of the transcriptional start site ( Fig . 4A ) . When the mouse CRM-GFP construct was introduced into zebrafish embryos , it drove expression in a pattern ( Fig . 4B ) very similar to that of the fish msxB gene as well as that observed endogenously in mice . These results suggest that both the zebrafish and mouse CRMs contain sufficient information to correctly direct expression to the dorsal ectoderm despite the fact that they show only limited sequence conservation . These observations provide another clear example of the highly conserved function of vertebrate CRMs from lineages that diverged over 400 MYA in the absence of obvious sequence conservation in these non-coding regions [33] , [34] . We pared down the zebrafish msxB CRM in transient transformant embryos and identified a minimal 671 bp fragment containing the most conserved island that also faithfully recapitulates msxB expression in dorsal neuroectodermal/neural crest progenitor cells ( Fig . 4D ) . Paralleling our approach in Drosophila , we searched for BMP responsive sites within the minimal msxB CRM by first scanning bioinformatically for candidate SE or AE sites using the SMAD1/5/8 consensus GNCKNC and SMAD4 consensus GNC ( T/V ) with relaxed spacing constraints , and then testing by EMSA whether oligonucleotides containing these sites could indeed assemble Drosophila pMM and/or pMMS complexes in response to BMP signaling in vitro ( Fig . S3 ) . This analysis identified a single highly conserved site ( zAE ) to which BMP signal-dependent pMM ( but no pMMS ) DNA binding was observed . The zAE contains candidate SMAD1/5/8 and SMAD4 binding sites separated by an unusually long 16 bp spacer ( Fig . S3A , B ) . These sites are also present in mouse albeit with different spacing ( 12 bp ) . Further analysis of this binding motif revealed that the SMAD1/5/8 and SMAD4 sites are each required , suggesting that the functional zAE includes both sites ( Fig . S3C ) . Changing the sequence or length of the spacer DNA linking the two sites did not affect the ability to form pMM complexes in vitro indicating that the exceptional length of the zAE spacer is not required for SMAD complex formation in vitro . Interestingly , however , changing the linker length to 5 bp allowed the formation of trimeric pMMS complexes ( Fig . S3D ) . We generated a 36 base pair deletion spanning the zAE ( and both the SMAD1/5/8 and SMAD4 candidate binding sites – DEL mutant ) in the context of the 671 bp msxB CRM and observed that GFP reporter gene expression was lost in transient transformant embryos ( Fig . 4D ) . Similarly , mutation of two core base pairs in the GC-rich region of the zAE ( GCR1 mutant ) , which abolished pMM binding in vitro ( Fig . 4C; Fig . S3C ) , also reduced reporter expression in vivo in transient transformant embryos ( Fig . 4D ) . These results indicate that a single BMP responsive site within the 671 bp zebrafish msxB CRM is required for mediating reporter gene activation by this element in vivo . The above dissection of BMP-responsive sequences within the Drosophila and zebrafish msh/msx CRMs suggests that they are under opposing forms of BMP regulation: repression in Drosophila versus activation in zebrafish . To test this hypothesis further , we compared the response of these CRMs to alterations in BMP signaling in vivo ( Fig . 5 ) . In Drosophila , we examined msh and ME reporter gene expression in both a dpp- mutant background and in embryos ectopically expressing dpp in the dorsal epidermis . As mentioned above , msh and ME-reporter gene expression both expand dorsally in dpp- mutant ( Fig . 1B; Fig . 5A ) . Conversely , ectopic dpp expressed from a Heat Shock-dpp construct ( HS-dpp ) resulted in loss of msh expression within its normal domain ( Fig . 5C ) . In zebrafish , a stable transgenic line carrying the 2 . 4 kb msxB-GFP reporter construct was crossed to lines carrying either a Heat Shock-chordin ( HS-CHD ) or a Heat Shock-BMP ( HS-BMP ) construct . When the BMP antagonist Chordin was induced by heat treatment ( Fig . 5G ) , msxB-GFP reporter expression was strongly suppressed , as was endogenous msxB expression ( Fig . 5D ) . The opposite effect was observed in HS-BMP embryos , however , where expression of endogenous ( Fig . 5F ) and reporter ( Fig . 5I ) genes was broadened compared to control embryos ( Fig . 5E and 5H , respectively ) that were subjected to the same conditions . Thus , consistent with the inverse effects of mutagenizing BMP-responsive sites in the Drosophila msh and zebrafish msxB CRMs , these two elements respond in an opposing fashion to equivalent manipulations of BMP signaling in vivo . Our analysis strongly suggests that BMPs pattern the neuroectoderm primarily via repression in Drosophila , while in zebrafish , BMPs function , at least in part , to activate the orthologous msxB gene .
Mutational analysis of the Drosophila msh CRM in this study supports a direct role for BMP repression acting via the SE1 and SE2 sites to suppress activity of this element in the dorsal ectoderm where there are likely to be moderate levels of BMP signaling . Mutation of either of these sites results in modest dorsal expansion of reporter gene expression while elimination of both sites by point mutations or a deletion spanning both sites causes more pronounced ectopic dorsal expression . The dorsal expansion in SE1 , SE2 double mutants is less complete , however , than that observed when the intact ME is crossed into a dpp- background , indicating that additional inputs are also involved in repressing the activity of this element dorsally . These additional BMP-dependent inputs might act either directly or indirectly . Since each of the three deletions spanning the remaining portions of the CRM ( i . e . sequences outside of the deletion covering the SE1 and SE2 sites ) result in reduced CRM activity it is possible that the effects of such hypothetical additional BMP responsive sites are canceled out by the deletion of necessary adjacent activation sites ( e . g . , deletion of the A2 site in the D3 region , Fig . 3C ) . If these hypothetical repressor sites act directly on the msh CRM they would presumably bind Mad , Medea and Schnurri , MAPK pathway transcriptional effectors , or possibly yet unknown BMP mediators alone or in conjunction with other transacting factors . Our detailed bioinformatic analysis and systematic experimental EMSA surveys have failed to identify any such sites , however . It is also possible that part of the BMP response of the msh CRM is mediated indirectly . For example , we have previously reported that localized overexpression of Brk can de-repress msh expression dorsally [7] , yet there are no consensus Brk sites in the ME and we were unable to detect Brk protein binding to any closely related candidate Brk sites by EMSA ( Fig . S1 ) . Thus , Brk may act via regulating expression of other components required for BMP signaling such as the BMP type-1 receptor Thick veins [35] . Alternatively , activators of the ME may be under negative BMP/Brk regulation . The SE1 and SE2 sites that play a role in repressing ME activity dorsally are imperfect matches to the consensus SE sites determined by Pyrowolakis and colleagues [36] . The ind CRM , however , which according to genetic data is more sensitive to BMP repression than msh [7] , contains a perfect SE site required for repressing activity of this element dorsally [20] . When the SE2 site in the ME was mutated to similarly match the ideal SE consensus sequence ( SE2* ) it repressed ME expression in its normal dorsal ectodermal domain in a dpp-dependent fashion ( Fig . 2C ) . In addition , competition experiments indicate that Mad/Schnurri/Medea bind to the ind-like SE element with higher affinity than the msh SE2 element ( Fig . S1D ) . These combined findings suggest that differences in affinity of SE sites for forming Mad/Med/Shn complexes contribute to the distinct responses of the two CRMs to BMP-mediated threshold-dependent repression . Using a combination of bioinformatics and efficient transgenesis in zebrafish we identified genomic fragments upstream of the zebrafish msxB and mouse Msx1 genes that drive neuroectodermal GFP-reporter gene expression at the open neural plate stage in zebrafish embryos . Further analysis of a minimal 671 bp zebrafish CRM identified a single conserved SMAD binding site that is required for activity of this element . An novel feature of this BMP-activation site is that the SMAD1/5/8 and SMAD4 binding site motifs are separated by a 16 bp spacer , which interposes approximately one and half turns of the DNA helix between these two sites , thus differing from other characterized vertebrate BMP activation sites in which these SMAD binding sites are closer [37] . Interestingly , deletion of 11 bp ( about one turn of the helix ) endows this modified site with the ability to bind the pMMS repressor complex in vitro . Whether this unique architecture of the msxB BMP activation site is relevant to activity within the neuroectoderm remains to be explored . We also examined the in vivo response of the endogenous zebrafish msxB gene and the msxB-CRM to inhibition of BMP signaling or ectopic expression of BMPs and compared these responses to equivalent manipulations of BMP signaling in Drosophila . In Drosophila , msh or ME-lacZ expression expands dorsally in a dpp- mutant while msh expression is repressed within its normal dorsal neuroectodermal domain by ectopic dpp expression . In contrast , expression of the zebrafish msxB gene , which is mirrored by activity of the msxB-CRM , is lost upon inhibition of BMP signaling and expanded or elevated in response to ectopic BMPs . Thus , both mutational analysis and in vivo testing suggest opposing mechanisms for BMP-dependent regulation of the msh and msxB genes in the early neuroectoderm . Given the opposing mechanisms by which the msh and msxB CRMs respond to BMPs , it is intriguing that a site within the msh CRM closely resembling an activation site ( AE2 ) is required for activation of this CRM . Also , another AE-like site ( AE1 ) lies within a region which when deleted greatly reduces ME driven reporter gene expression , although the role of that AE1 site remains to be examined . These AE-like sites , while having only single mismatches to consensus Mad-Medea binding sites , did not bind Mad-Medea complexes in vitro , indicating that they are most likely not involved in mediating a BMP response . Additionally , experiments designed to identify potential positive roles of BMP signaling in regulating ME activity provided no evidence for such an effect . Given the known role of AE sites in other genes to BMP-dependent activation and the evidence that BMPs can act positively to promote msx gene expression in vertebrates , it is tempting to speculate that these sites could once have been BMP responsive activation sites and were subsequently co-opted by different transcription factors ( possibly a TAGteam motif [38] binding protein ) in the course of evolution to maintain msh expression in a BMP-independent fashion . Identifying such transcriptional activators is an interesting goal for future experiments . In Drosophila , Evo/Devo studies of the even-skipped stripe 2 CRM have suggested that regulatory mechanisms that lead to a particular gene expression pattern are extremely flexible , i . e . , the same pattern can be achieved in multiple ways [39] . Accordingly , in the current case of BMP-dependent regulation of msh/msxB expression , natural selection may have operated similarly to maintain relevant gene expression patterns that fulfill a particular function ( i . e . dorsal neuroectodermal expression ) while allowing the upstream mechanisms generating that pattern to change over time . As summarized above , our analysis strongly suggests that BMPs pattern the neuroectoderm primarily via repression in Drosophila , while in zebrafish , BMPs function , at least in part , to activate the orthologous msxB gene . Genetic studies and exogenous BMP treatment in zebrafish suggest that msx gene expression may also be repressed by high levels of BMP signaling . Whether the BMP-responsive site in the 671 bp msxB CRM together with other potential BMP-responsive elements mediate such a biphasic response will be interesting to address in future experiments . In the future , it will also be important to determine whether expression of other msx paralogs in the dorsal CNS of zebrafish ( e . g . , msxC , E [31] ) or msx genes in other vertebrates ( e . g . , the murine Msx1 neuroectodermal CRM identified here ) are similarly regulated by BMPs . Analysis of these additional vertebrate msx CRMs should reveal whether distinct evolutionary trajectories have shaped the BMP responsiveness of these elements . Such comparative studies may also shed light on whether there is a single or multiple independent origin ( s ) of BMP regulation of vertebrate msx genes . Furthermore , analysis of the CRM driving BMP-dependent expression of an echinoderm Msx homolog in regions of peak BMP activity [14] will be informative since this gene is expressed in the non-neural ectoderm . In this case , one might predict finding only positively acting AE-like BMP-responsive sites . There are two possible explanations for distinct mechanisms of BMP-regulation of msh/msxB expression in flies versus fish . One is that these genes independently evolved BMP responsiveness . Alternatively , BMP-dependent regulation may be an ancestral trait dating back to the first bilaterians with a condensed CNS . We favor the latter alternative for the following reasons . First , the co-linearity of msh-msx , ind-gsh/pax , and vnd-Nkx2 . 2 genes relative to the source of BMPs and the BMP responsiveness of these genes in species from all three primary branches of bilateria - flies ( ecdysozoa ) , vertebrates ( deuterostome chordates ) , and annelid worms [40] ( lophotrochozoa ) - provides a compelling argument for this arrangement reflecting the ancestral state . Second , a polarized source of BMPs was present in diploblasts ( e . g . , corals [41] , [42] , jellyfish [43] , and the sea anemone [44] , [45] ) and therefore preceded evolution of bilaterian triploblasts and a condensed CNS . Thus , it is plausible that a single species evolved a condensed CNS which deployed neural identity genes along the DV axis in much the same way that Hox genes are expressed in sequential order along the AP axis . Finally , if one looks more broadly among the 30 bilaterian phyla , a striking trend is that at least some clades within most of these phyla have a condensed CNS with three primary axon bundles [46] , suggestive of an ancestral tripartite subdivision of the CNS . It is true that there are also many examples of species scattered among these phyla that either secondarily lost a condensed polarized CNS or retained a prior ancestral state in which there was only a distributed nervous system . Echinoderms in which Msx genes are expressed in the non-neural ectoderm ( see above ) or the hemichordate Saccoglossus kowalevskii which has lost bilateral symmetry to become radially organized [47] may be examples of such derived simplifications of the nervous system . Thus , in our view , the most likely scenario is that the ancestral bilaterian CNS was a condensed nervous system partitioned into at least three DV domains and that loss of centralization has occurred numerous times in different lineages undergoing morphological simplification . If one assumes a common ancestral origin for BMP-regulation of msx genes , one can imagine various scenarios under which BMP-mediated regulation of msh/msx genes could have switched its effect during evolution . In vertebrates , BMP targets frequently contain Drosophila SEs that activate rather than repress transcription . This might be due to Shn proteins losing their repressive activity through changes in the Shn amino acid sequence and/or the lack of components required for repression downstream of Shn . The molecular relatedness of SEs and AEs raises the possibility that ancestral SE-mediated repressive effects on msh/Msx expression may have been relatively easy to convert into activating effects in the vertebrate lineage by the loss of the Shn repressor function . Consequently , the increased linker length of zAE could be accounted for by the lack of evolutionary pressure on the SE to meet the sequence requirements for Shn recruitment . Since the Drosophila msh gene is weakly repressed by BMPs ( e . g . , relative to ind and other neural genes such as AS-C , scrt or sna [6] ) , while vertebrate msx genes are weakly activated by BMPs ( i . e . , high neuroectodermal levels of BMPs are required to activate msx genes ) an intermediate CRM state may have existed in which BMPs both weakly activated msx gene expression within the neuroectoderm at moderate levels while repressing gene expression at the peak BMP levels present in the adjacent epidermis . Indeed the zebrafish msxB gene may represent such a bifunctional intermediate condition since in vivo studies indicate that high levels of BMPs can inhibit msxB expression [12] . It remains to be determined whether such proposed positive and negative inputs are mediated by a single or multiple independent CRM ( s ) . Within different evolutionary lineages such biphasic responses could have then been rendered monophasic in opposing directions to account for the observed differences in the Drosophila versus vertebrate or echinoderm Msx CRMs . In vertebrates , one potential driving force for reducing the effect of BMP-mediated inhibition may have been the incorporation of BMP expression within the dorsal neural tube itself since this would be expected to generate much higher BMP levels than would result from BMPs diffusing in from the adjacent epidermal ectoderm ( e . g . , as is the case in Drosophila ) . In future analyses it will also be important to examine BMP-mediated regulation of additional neural identity genes expressed along the dorsal-ventral axis including the Gsh ≈ ind and Nxk2 . 2 ≈ vnd genes as CRMs controlling expression of each of these genes will have undergone independent evolutionary trajectories . Since there is evidence that laterally and ventrally expressed genes in vertebrates are inhibited by BMPs [48]–[52] , and because the more ventrally expressed ind gene in Drosophila is more sensitive to BMP-mediated repression than msh [7] , one might expect to find similar , and perhaps conserved ancestral modes , of BMP-mediated repression of these genes across bilateria . It will also be interesting to understand how flexible the ancestral metazoan state was by investigating the relationship between BMPs and msx genes in basal metazoans such as jellyfish . In these diploblastic animals , although the BMP-msx relationship has not been tested , BMP2/4 [53] and msx [54] homologues are expressed in adjacent regions during development , as is the case in the majority of triploblastic animals .
We identified candidate SE and AE sites in the msh , msxB and msx1 CRMs using binding site consensus sequences curated from the literature referenced and used Gene Palette [55] . For this analysis , we used the consensus sequence GNCGNC ( N ) 5GNCTG to identify candidate Silencer Elements ( SE ) and the consensus GGCGCCA ( N ) 4GNCV for Activator elements ( AE ) allowing for single base-pair mismatches to these consensus sequences . We identified candidate zebrafish msxB and mouse Msx1 CRMs by using genome wide alignments for multiple vertebrate species , which indicates regions of high sequence conservation as provided by the UCSC genome browser ( http://genome . ucsc . edu ) . The 700 bp msh CRM is described in Von Ohlen et al . , 2009 [25] . All primers used in this study and the corresponding constructs generated can be found in Table S1 . The various Drosophila msh-CRM constructs were subcloned in pCR-TOPO vectors ( Invitrogen ) and subsequently cloned into the [P]acman vector [28] as NotI and KpnI restriction fragments . Site-directed mutagenesis PCR methods were adapted from [56] . The primers used to isolate the zebrafish msxB CRMs and the mouse msx1 CRM can be found in Table S1 . Zebrafish constructs were cloned into pENTR-TOPO ( Invitrogen ) , transferred to pTol2 by Gateway Recombination and injected in zebrafish embryos as previously described [30] . The Drosophila dpph46 null allele used in this study is Flybase stock number 2061 . The 8x HS-dpp stock and its use are described in Biehs et al 1996 [6] . The schnurri mutant allele is shn04738 . To generate the dl dpp st2-dpp+ embryos , females that are Dpdpp/+; dl1 cn1 sca1/dpph46 wgsp dl1 were crossed to yw/Y; dpph46 wgsp st2-dpp+ , w+/CyO males . The fly strain used to inject all constructs has genotype PBac{yellow[+]-attP-3BVK00002 and injections were outsourced to BestGeneInc ( http://www . thebestgene . com/ ) . The zebrafish strains containing the hsp70:bmp2b [57] and hsp70:chd [58] transgenes were crossed to stable transgenic lines containing the msxB-CRM construct . Embryos at the sphere stage ( 4hpf ) were subjected to heat shock at 37°C for 1 hour and then returned to normal temperature of 28 . 5°C until they were fixed at the bud – 6 somite ( 10–11hpf ) stage as necessary . For electrophoretic mobility shift assays ( EMSA ) , Drosophila S2 cells were co-transfected with 50 ng TkvQD and 175 ng Mad- and Med-expression plasmids or with 400 ng of a ShnCT-expression plasmid . Cells were harvested 72 hr after transfection and lysed for 10 min at 4°C in 100 µl of 100 mM Tris ( pH 7 . 5 ) , 1 mM DTT , 0 . 5% TritonX100 and 1%NP40 . Radioactively labeled probes were generated by annealing and filling in partially overlapping oligonucleotides in the presence of [P]-32 ATP . Binding reactions were carried out in a total volume of 25 µl containing 12 . 5 µl 2x binding buffer ( 200 mM KCl , 40 mM HEPES ( pH 7 . 9 ) , 40% glycerol , 2 mM DTT , 0 . 6% BSA and 0 . 02% NP40 ) , 10000 cpm of radioactively-labeled probe , 1 µl poly dIdC ( 1 mM ) and 7 µl of cleared S2 cell extracts . After incubation for 30 min at 4°C , the reactions were analyzed by non-denaturing 4%polyacrylamide gel electrophoresis followed by autoradiography . Fluorescent in situ hybridization methods used were performed according to [59] in Drosophila embryos and adapted to zebrafish embryos by increasing the hybridization temperature: 55°C in Drosophila to 65°C in zebrafish embryos . Antibodies used: dpERK ( Cell Signaling Technology #5683 ) , anti-digoxigenin ( Roche ) , anti-biotin ( Roche ) , Alexa fluor 488 , 594 , 647 ( Invitrogen ) . We also used colorimetric staining methods performed according to O'Neill and Bier [60] . The DNA template used to generate the msxB probe was a generous gift from the Riley lab . Histochemical stain images were acquired using a Nikon optical microscope and fluorescent stain images were collected using a LEICA SP2 confocal microscope . Images were adjusted for color , brightness and contrast using Adobe Photoshop software .
|
The trunk nervous system of both vertebrates and invertebrates develops from three primary rows of neural stem cells whose fate is determined by neural identity genes expressed in an evolutionarily conserved dorso-ventral pattern . Establishment of this pattern requires a shared signaling pathway in both groups of animals . Previous studies suggested that a shared signaling pathway functions in opposite ways in vertebrates and invertebrates , despite the final patterning outcomes having remained the same . Here , we employ bioinformatics , biochemistry , and transgenic animal technology to elucidate the genetic mechanism by which this pathway can engage the same components to generate opposite instructions and yet arrive at similar outcomes in patterning of the nervous system . Our findings highlight how natural selection can act to conserve a particular output pattern despite changes during evolution in the genetic mechanisms underlying the formation of this pattern .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"genetics",
"genome",
"evolution",
"gene",
"regulation",
"neuroscience",
"developmental",
"biology",
"stem",
"cells",
"molecular",
"genetics",
"developmental",
"neuroscience",
"animal",
"cells",
"gene",
"expression",
"molecular",
"evolution",
"neural",
"crest",
"evolutionary",
"genetics",
"molecular",
"biology",
"invertebrate",
"genetics",
"neural",
"stem",
"cells",
"cell",
"biology",
"gene",
"regulatory",
"networks",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"computational",
"biology",
"evolutionary",
"biology",
"cell",
"fate",
"determination",
"evolutionary",
"developmental",
"biology"
] |
2014
|
BMPs Regulate msx Gene Expression in the Dorsal Neuroectoderm of Drosophila and Vertebrates by Distinct Mechanisms
|
Due to its inhibition of the Abl kinase domain in the BCR-ABL fusion protein , imatinib is strikingly effective in the initial stage of chronic myeloid leukemia with more than 90% of the patients showing complete remission . However , as in the case of most targeted anti-cancer therapies , the emergence of drug resistance is a serious concern . Several drug-resistant mutations affecting the catalytic domain of Abl and other tyrosine kinases are now known . But , despite their importance and the adverse effect that they have on the prognosis of the cancer patients harboring them , the molecular mechanism of these mutations is still debated . Here by using long molecular dynamics simulations and large-scale free energy calculations complemented by in vitro mutagenesis and microcalorimetry experiments , we model the effect of several widespread drug-resistant mutations of Abl . By comparing the conformational free energy landscape of the mutants with those of the wild-type tyrosine kinases we clarify their mode of action . It involves significant and complex changes in the inactive-to-active dynamics and entropy/enthalpy balance of two functional elements: the activation-loop and the conserved DFG motif . What is more the T315I gatekeeper mutant has a significant impact on the binding mechanism itself and on the binding kinetics .
The revolutionary discovery of the potent anticancer drug imatinib ( Gleevec , 2001 ) [1] had a huge impact on cancer therapy . This drug has a striking efficacy in the early stages of chronic myeloid leukemia ( CML ) , with 90% of patients showing remission [2 , 3] . Imatinib targets the Abl tyrosine kinase ( TK ) , constitutively active in CML due to a chromosomal translocation [4] . Unfortunately , most patients in an advanced stage of the disease suffer from relapse due to the onset of drug-resistance [5] . Even if , next-generation kinase inhibitors ( KIs ) are available , or in clinical trials [6] , their efficacy might also be affected by drug resistance responses . Among different mechanisms , the development of resistance-inducing mutations is the most relevant in tyrosine kinases [6] . Mutations occur in highly conserved positions on the protein [7] , frequently shared by several kinases [8] , suggesting a conserved kinome-wide mechanism . Unfortunately , the molecular mechanism of mutation-mediated resistance are only partially understood . In the case of the widely studied “gatekeeper” mutant , found in several TKs ( T315I in Abl ) [9] , three mechanisms have been proposed . The direct one involves the abrogation of a crucial hydrogen bond formed by imatinib . A second hypothesis posits that the observed shift towards the active form , which was reported in Abl and several other TK bearing the gatekeeper mutation , would allow the natural substrate ATP to outcompete the inhibitors . [10–13] Very recently , a third mechanism has been proposed for Abl T315I whereby the suppression of an induced fit effect involving the p-loop would be responsible for the decreased binding affinity of imatinib . [14] It is probable that the gate-keeper mutations have a combined effect on the binding of inhibitors , changing their binding mode and affecting at the same time the conformational changes [10 , 11] . The importance of the conformational changes in the mode of action of drug-resistant mutations [15 , 16] is also confirmed by the fact that many of them are far away from the binding site ( Fig 1 ) , and thus act allosterically by disfavoring the drug-binding conformation and favoring active form [8 , 17–19] . The link between conformational changes and allosteric regulation in TKs is well established . For instance , in the case of Src ( a close homologue of Abl ) the gatekeeper mutation has been shown to allosterically affect remote regulatory motifs [20] . Indeed , TKs can exist in a dynamic equilibrium between multiple conformations [21–23] , differing by the conformation of the “activation loop” ( A-loop ) , of the conserved DFG motif and of the αC-helix ( Fig 1 ) . While the features of the active , catalytically-competent , state are shared among different TKs , and include an A-loop in an extended ( “open” ) conformation , inactive conformations can be multiple and highly diverse , all sharing an ( at least partially ) “closed” A-loop . Type II inhibitors , as imatinib , target a particular inactive conformation , known as “DFG-out” , where the aspartate is flipped , taking the place of the phenylalanine and pointing out of the ATP binding site [24] and opening an adjacent “allosteric” pocket . When both the DFG is in the ‘out’ conformation and the A-loop is not fully extended , the drug can enter the cavity and adopt a bridge position above the DFG , occupying both the catalytic site and the allosteric pocket [25] . The DFG-out conformation has been observed in many kinases but , despite a common binding mode , imatinib binds strongly only to some of them [26] . In contrast to the induced-fit mechanism proposed for Abl , free energy calculations performed with different and independent approaches on several TKs are consistent with a conformational selection mechanism [23 , 27 , 28] . Indeed , the different affinity to Abl and Src can be explained by the thermodynamic penalty to adopt the drug-binding “DFG-out” conformation [29] . Our simulations on Abl and Src also revealed a correlation between the flexibility of specific functionally-relevant structural elements and the DFG-out penalty [29] . To investigate the molecular mechanism of drug-resistant mutations , and whether or not changes in the conformational landscape , have a role in it , here we performed enhanced-sampling atomistic molecular dynamics ( MD ) simulations and free energy calculations on a pool of wild-type ( WT ) TKs for which imatinib has various strengths ( IC50 ) , and several Abl Imatinib-resistant mutants . In most drug resistant mutants we find extensive changes in the conformational free energy landscapes associated with two functionally-important conformational changes: the DFG-flip and the closed to open A-loop switch . The changes between the relative free energies of the active ( DFG-in , extended A-loop ) and inactive ( DFG-out , closed A-loop ) states are in the main due to entropic contributions arising from the fast ( sub-μs ) dynamics of proximal structural elements . It is thus not surprising that we find a correlation between the sub-μs flexibility of specific structural elements in the active state and drug resistance . Indeed , altering the sub-μs dynamics has an effect on the binding of Imatinib , as also shown by mutagenesis and calorimetry experiments . The gatekeeper mutant T315I is a notable exception , as it affects both the binding mechanism itself and the conformation of the A-loop equilibrium . To the best of our knowledge this is the most in-depth computational study addressing the molecular roots of resistant mutants of Abl .
The kinase structures were retrieved from the Protein Data Bank ( PDB entries 2G1T , 2SRC , 1PKG , 3KMM , 2WGJ and 2ITW ) . Missing residues were added using the software Modeller [30] , according to the respective Uniprot sequences . We used the Amber99SB*-ILDN [31 , 32] force field , including backbone corrections by Hummer and Best [33] , with explicit TIP3P [34] water molecules . The unbiased MD simulations were carried out with the ACEMD program [35] running on GPUs . The systems were minimized with 10000 steps of conjugated gradient and equilibrated in the isothermal-isobaric ( NPT ) ensemble for 10 ns , using a Berendsen barostat at 1 atm . The temperature was kept at 305 K by a Langevin thermostat . A 400 ns production run was carried out for all the systems in the canonical ( NVT ) ensemble . The runs for Abl , Src and Kit were extended to a total length of 1 μs in both the DFG-in and DFG-out conformations . The simulations of the five Abl mutants ( G250E , E279K , H396P , E450K and T315I ) were carried out with the same setup in both the DFG-in and DFG-out conformation , after mutating in-silico the respective residue in the Abl structure . To capture the sub-μs dynamics , the RMSF of the Cα were averaged on 30ns non-overlapping windows after discarding the first 100ns of each run . We also compared the diffusion in the space described by the first two principal component analysis vectors ( see S9 Fig ) . Since the conformational changes of TKs take place on time scales longer than those accessible by standard MD simulations , here we used a combination of enhanced sampling approaches . Parallel Tempering—Well Tempered Metadynamics [36 , 37] in the well tempered ensemble ( WTE ) [38 , 39] ( PTmetaD ) was chosen due to its proven ability to fully converge complex conformational free energy surfaces such as those relevant in kinases ( including the DFG-flip ) . Indeed , the PTmetaD approach ( both in the standard and WTE variants ) has already successfully used to study the conformational dynamics and its associated free energy landscape in many kinases [12 , 13 , 29 , 40 , 41] and other flexible proteins [39 , 42] . PTmetaD was performed using the software Gromacs 4 . 5 [43] and the PLUMED plugin [44] , using an integration step of 2 fs . The particle mesh Ewald algorithm was used for electrostatic interactions . Temperature coupling was done with the V-rescale algorithm [45] . The WTE allowed the use of a reduced number of replicas compared to standard PTMetaD [12 , 39] . An average exchange probability of 24% was obtained using 5 replicas in the temperature range 305–400 K . We used the same four collective variables ( CVs ) that were used to reconstruct the free energy surface ( FES ) associated with the DFG flip in Src and Abl [29] . They are shown in S6 Fig and defined in the following ( SRC numbering ) : CV1 is the distance between the centre of mass of Asp404 ( DFG motif ) and Lys295 . CV2 is the distance between the centre of mass of Phe405 and the Cβ of Ile293 . CV3 , is a function of 3 dihedral angles f ( ϕ404 , ψ405 , ψ408 ) ranging from 3 , when the three dihedral arguments correspond to the DFG-Asp-in position , to 0 , when they are in the DFG-Asp-out conformation . CV4 is the distance between the centre of mass of residues Asn381 , … , His384 and residues Ala408 , … , Ile411 of the activation loop , known to form a β-sheet in the active conformation . The height of the Gaussians was set at 2 . 0 kJ/mol with a deposition rate of 1/2000 steps and a bias factor of 5 . The Gaussian width used for the CVs was 0 . 1 for the dihedral similarity ( CV3 ) and 0 . 3 Å for all the others . A minimum of 400 ns of sampling per replica in the NVT ensemble were needed to reach full convergence of the free energy . In the case of G250E and T315I , as the convergence was slower than in the other cases , we performed more than 600 ns and 1200 ns of sampling per replica , respectively . The total sampling time amounted to more than 14 μs across all mutants . The free energy surface reconstruction was obtained from the PT-metaD-WTE by reweighting the fixed potential energy bias and using two independent approaches: integrating the bias or using the time independent estimator of Ref . [46] . The convergence of the free energy reconstruction was monitored by integrating the cumulative added bias as a function of time ( 50 ns intervals ) and comparing the reconstruction to that obtained by the time-independent estimator ( as shown in S11 , S14–S18 Figs ) . Changes of all the CVs used were also monitored to guarantee that the system diffuses freely in the CV space and is able to visit all the basins several times ( see S12 Fig ) . The FES were also reprojected as a function of two other CVs describing the conformational change of the A-loop from open to closed . Two path collective variables [47] were built by using the open and closed crystallographic structures of Src ( CV1 ) and Abl ( CV2 ) . The reweight was performed by using our python implementation of the approach of Ref . [46] ( available on our homepage: https://www . ucl . ac . uk/chemistry/research/group_pages/prot_dynamics/ . The entropic contribution to the DFG-in DFG-out free energy difference was computed by linear fitting of the free energy differences as a function of temperature . It has been shown that this approach is more accurate than competing ones . [48] . We also tested the robustness of the estimate with respect to the range of temperatures on both standard PT and WTE-PT ( see S8 Fig ) . The drug binding free energy surface was calculated using metadynamics with the same software described above . The General Amber Force Field ( GAFF ) was used for the ligand . The ligand charges were calculated at the HF level using a 631-G* basis set with the Gaussian03 [49] package . QM-level torsional scans with a step of 10 degrees were carried out for the ca-ca-n-c and ca-ca-c3-n3 dihedrals which appeared to have wrong torsional profiles . The profiles with the GAFF force field in vacuum were then fitted to the QM ones to refine the dihedral parameters . As customary in the case of ligand binding [50 , 51] , we performed a preliminary metadynamics run using sub-optimal geometrical CV to obtain an initial pathway for the setup of the path collective variables ( PCV ) [47] . We selected 23 frames from the lowest free energy path obtained in the preliminary run and optimized this initial guess using the methodology described by Branduardi et al . [47] . To take into account possible rearrangements , we included the Cα atoms of the A-loop and of the αC-helix in the definition of the PCVs . The 2 PCVs s and z were used to run a 300 ns metadynamics . The height of the Gaussians was set at 2 . 0 kJ/mol with a deposition rate of 1/2000 steps and a bias factor of 10 . The Gaussian width used for the CVs was 0 . 1 for s and 0 . 003 for z . The free energy corresponding to the last leg of the unbinding mechanism ( from the external binding pose to a fully solvated state ) was computed again by using Well-tempered metadynamics following the approach of Ref . [52] . Src was expressed and purified following the procedure detailed in Ref . [53] . For the first two mutants , an auto-induction protocol was used to reach a good bacterial expression . Thermodynamic binding parameters for the association of imatinib to Src wild type and to the designed mutants were acquired using a VP-ITC microcalorimeter ( MicroCal Inc . ) . The sample solution consisted of 2 ml 0 . 1 mM protein in MES buffer ( pH 6 . 5 ) supplemented with 2% DMSO . The ligand solution consisted of 2 ml 1 . 6 mM imatinib in MES buffer ( pH 6 . 5 ) with 2% DMSO . Both solutions were degassed for 5 minutes under vacuum . Titrations were conducted at 30°C , consisted of a first control injection of 1 μl followed by 37 injections of 8 μl over 16s , each spaced by an equilibration period of 480s . The samples were stirred at 266 rpm . Raw data were collected and corrected for ligand heats of dilution . A one site binding model was assumed and data were fit using MicroCal Origin software ( version 7 . 0 ) . All experiments were repeated in triplicate .
To understand whether inducing a change in the flexibility of relevant structural motifs is sufficient to increase the drug sensitivity , we designed a number of Src KD mutants and measured the changes in Kd . We engineered mutations in four different locations , the P-loop ( Q275 ) , the β3-αC loop ( P299 ) the A-loop ( Q420 ) and the αG-helix ( V461 ) , to the equivalent residues in Abl and Lck . The residues to be substituted were chosen in the functionally-relevant regions showing a major variation in the sub-μs dynamics , between the weak and strong imatinib binders . We first performed MD simulations and compared the flexibility of various mutants to that of Src WT . Based on the RMSF we choose three candidates for the experiments . The first bears Q275A , P299Q and Q420E substitutions ( Src to Lck ) , the second Q275G , P299E and Q420A ( Src to Abl ) and a third one with Q275A , P299Q and V461S ( Src to Abl ) . The RMSF shows an increased flexibility in the N-lobe ( p-loop , β3-αC loop and αC helix ) compared to Src WT ( see S3 Fig ) . For the mutant Q275G/P299E/Q420A , also an increase in A-loop flexibility was identified . In the Q275G/P299E/V461S mutant was introduced a mutation on the αG-helix , that has provoked a suppression of the fluctuations in this area , resembling the αG-helix of Abl . The Kd appear to be anti-correlated with the RMSF of the A-loop ( Table 2 ) . As a cross-validation , when a chimeric protein with the entire αC-helix of Abl was produced , we observed a general suppression of the dynamics and an increased Kd of 14 μM . At difference with previous studies reporting similar Kd [26] , the mutated residues are far from the binding site and do not interact with the drug . If the correlation between changes in the sub-μs dynamics and imatinib affinity observed in the MD simulations and validated by experiments is due to entropic effects it must be reflected in the conformational free energy landscape ( and in the entropic contribution to the relative stabilities of the DFG-in and -out states ) . As discussed above , Imatinib binds to an inactive state in which both the conformation of the DFG motif and that of the A-loop are important . In non-phosphorylated TKs the active-like “open” state in which the A-loop is extended , is marginally populated [12 , 63 , 64] . The DFG-out inactive state has a significant thermodynamic penalty and ( according to the observed changes in the sub-μs dynamics ) is in most cases entropically disfavored . In both inactive states the delicate balance between entropic and enthalpic contributions are crucial in defining their stability with respect to the open , DFG-in active state . This balance can be easily upset by drug-resistant mutations either directly or through the allosteric network . Thus a mutation affecting the entropy / enthalpy balance of the active and inactive states could easily shift the kinase towards active state , decreasing the binding affinity of type II inhibitors . Indeed , the stabilization of the active ( extended A-loop ) conformation has been experimentally observed for many TK gatekeeper mutants ( including Abl T315I ) [10 , 11 , 13 , 65] . To further investigate this issue , we performed large-scale multiple-replica PT-MetaD simulations and computed the fully converged conformational free energy landscape associated with the DFG-flip of the five drug-resistant mutants . In Fig 3 we report the FES projected as a function of two collective variables ( CVs ) that distinguish the DFG-in from the DFG-out state [29] , namely the distance between the DFG Asp404 and Lys295 and between the DFG Phe405 and Ile293 ( Src numbering ) . As expected , for all mutants the global minimum of the FES corresponds to the DFG-in state ( basin “IN” in Fig 3 ) . The mutants also explore the DFG-out state ( basin “OUT” ) , which in most cases corresponds to a well-defined metastable minimum . All mutations alter significantly the conformational free energy landscape . The free energy penalty of the DFG-out state increases going from Abl WT and G250E ( ΔG of 4 ± 0 . 5 kcal/mol ) to E450K ( 6 kcal/mol ) , H396P and E279K ( 7 ± 0 . 5 and 8 ± 0 . 5 kcal/mol ) . The gatekeeper T315I and ( to a lesser extent ) G250E are significant exceptions as the ΔG associated to their DFG-flip is very close to that of the WT . The increased thermodynamic penalty is due to a net entropic loss of the DFG-out state , as shown in Table 3 ( the DFG-in/-out free energy differences as a function of the temperature are reported in S8 Fig ) . Again T315I and G250E are special cases as there is an entropic gain for the DFG-out state . The entropic contributions to the DFG flip are in excellent agreement with the observed changes in the sub-μs dynamics . Indeed in H396P and E279K , where the A-loop is significantly more rigid , the flip has a larger free energy penalty . In the cases of G250E and the T315I gatekeeper mutant the penalty for the flip is comparable to that of the WT , and in agreement with the observed increased flexibility of the N-lobe and A-loop in the DFG-out state , there is an entropic gain associated with it . In both cases the DFG-out has a peculiar geometry where both Asp404 and Phe405 point outwards ( S4 Fig ) and an open active-like extended A-loop conformation is observed . To quantify this aspect , we analyzed how the population of the “open” state is altered by the mutations . We re-projected the free energy surfaces ( FES ) as a function of path variables [47] describing the opening of the A-loop ( Fig 4 ) with respect to Src ( CV1 ) and Abl ( CV2 ) . Abl , Src and E279K mainly populate semi-closed and closed conformations ( basin B and C ) , while G250E , T315I and to a lesser extent E450K , H396P show a minimum in correspondence of the open A-loop ( basin A , CV1 < 6 ) . The A-loop in these and other mutants tends to form a second helix turn as in Src ( S5 Fig ) . The residues involved are Asp391 to Ala395 , which in the case of H396P are right before the mutation . The formation of this helix has been shown to stabilize the extended A-loop active state in TKs [10] and is also involved in the mechanism of oncogenic mutations [7 , 12] . The stabilization of the active state is known to weaken the binding of ATP-competitive inhibitors to oncogenic mutants of EGFR [10 , 12] and in the phosphorylated form of Abl [25 , 26] . In the case of T315I the observed stabilization of the active state is in agreement with multiple experimental observations [10 , 11 , 13 , 65] . Finally , we also observe a loss of structural integrity of the αC and αG helices in the DFG-out state of most mutants ( S4 Fig ) in agreement with previous proposals stating that conformational transitions in kinases are accompanied by local unfolding of secondary structural elements [66 , 67] . Our results appear to be consistent with a significant impact of the resistant mutations on two different conformational changes: the DFG flip , disfavoring the drug-binding conformation ( mainly mutants with rigidified A-loop and N-lobe ) , and the opening of the A-loop , possibly favoring the locking of the kinase in an active form . However , from the simulations on the unbound kinases , we cannot rule out an effect of the gatekeeper mutation on the proposed “induced-fit” mechanism [14] . Indeed , the predicted ΔG associated to the DFG-flip of T315I , which is very close to that of the WT , leaves two hypotheses open . Either the weaker binding of imatinib is due to the observed stabilization of the extended A-loop active state , or to the suppression of an induced fit effect , possibly acting on the p-loop conformation . To clarify this point and shed more light on the binding mechanism itself , we have computed the binding free energy of imatinib to Abl WT and Abl T315I along a physical association pathway . The use of such an approach , albeit it is significantly more expensive than an “end-point” free energy calculation ( e . g . thermodynamic integration ) has the advantage of reporting on free energy barriers and thus on the binding and unbinding kinetics . As expected , the crystallographic binding pose corresponds in both cases to the deepest minimum . Starting from that pose , the bidimensional ( un- ) binding free energy profiles show a substantial difference in imatinib’s mechanisms of binding to Abl WT and T315I ( Fig 5 ) . In the WT kinase the barrier to unbinding is lower ( in agreement with the observed differences in unbinding kinetics ) and there is one main exit path . It is also interesting to note that we find an “external binding pose” from which the inhibitor slides to its final crystallographic binding pose . In the external binding pose the DFG motif is already in the “out” position . In T315I we observe two different unbinding / binding pathways ( B’ and E , Fig 5 and S10 Fig ) somewhat similar to previous proposals . [68] In both of them the p-loop has a significant role . A pose similar to the external binding pose observed in the WT kinase is present ( C’ in Fig 5 ) but it is slightly shifted towards the αC helix and more stable than in the WT . Thus the gain in free energy from the external binding pose to the final ( crystallographic ) pose is lower , in agreement with the observed increase of the IC50 . When the unbinding free energy from C to a fully solvated state is accounted for ( see S13 Fig ) , the overall free energy difference from the unbound fully solvated state to the crystallographic pose is around 13 kcal ± 2 ( the larger convergence error is due to the algorithms used ) . When the free energy penalty of the DFG flip ( ≃ 4 kcal ) is subtracted , we get 9 kcal ± 2 , in agreement with the calorimetric and IC50 experimental data . It is thus clear that the effect of the T315I gatekeeper mutation is dual . It both affects the binding mechanism and stabilizes the active , extended A-loop conformation . We have studied the interplay between sub-μs dynamics and conformational dynamics in TKs , and how these are influenced by drug-resistant mutations impacting Type-II inhibitors binding . Overall , our simulations show that drug-resistant mutations have a significant effect on both the sub-μs dynamics and the conformational free energy landscape . They affect the energy and the population of the DFG-out inactive state and of the open A-loop active-like state . Since type II inhibitors bind to the inactive kinase , a more accessible DFG-out state leads to a stronger binding of imatinib and other type II inhibitors . On the contrary , a more populated active-like state , in which the A-loop is elongated , increases the affinity towards ATP and disfavors type II inhibitors . The selected drug-resistant mutants fall in two partially overlapping categories: those that have a significantly higher free energy penalty for the DFG-out state ( E279K , H396P ) and those ( G250E , E450K , T315I ) that populate the A-loop open active-like state . What is more the T315I gatekeeper mutant has a significant impact on the binding mechanism itself and on the binding kinetics . The mutations affect the free energy differences associated to the conformational changes mainly by changing the sub-μs dynamics and consequently the entropy / enthalpy balance of the different states . Thus , relative short MD simulations , by revealing changes in the sub-μs dynamics , might be used to predict the impact of new mutations on Imatinib resistance . The important role played by the entropic penalty of the DFG-out state must also be kept in mind when comparing low temperature and room temperature experiments . On the whole , we characterized the mechanism of action of several drug-resistant mutants of Abl . We have shown the link between fast and slow dynamics in these complex systems , providing a deeper understanding of the thermodynamics , kinetics and allosteric regulation of type II inhibitor binding TKs . In perspective , our results could help the design of fast and predictive computational approaches to predict the effect of yet unknown mutations of TKs .
|
Imatinib remains the most important and studied anti-cancer drug for cancer therapy in its new paradigm . Due to its inhibition of the Abl kinase domain , imatinib is strikingly effective in the initial stage of chronic myeloid leukemia with more than 90% of the patients showing a complete remission . However , the emergence of drug resistance is a serious concern . Here , we investigate the molecular mechanism of drug-resistant mutations which , despite the importance and the adverse effect on cancer patients prognosis , is still debated . Our extensive molecular simulations and free energy calculations are consistent with an allosteric effect of the single-point drug-resistance-causing mutations on the conformational dynamics . Two partially independent conformational changes play a role . Our findings might help the design of anti-cancer therapies to overcome drug resistance and be used to predict the clinical relevance of new drug-resistant mutants found by genetic screenings of tumor samples .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] |
[] |
2015
|
Towards a Molecular Understanding of the Link between Imatinib Resistance and Kinase Conformational Dynamics
|
Physical activity and molecular ageing presumably interact to precipitate musculoskeletal decline in humans with age . Herein , we have delineated molecular networks for these two major components of sarcopenic risk using multiple independent clinical cohorts . We generated genome-wide transcript profiles from individuals ( n = 44 ) who then undertook 20 weeks of supervised resistance-exercise training ( RET ) . Expectedly , our subjects exhibited a marked range of hypertrophic responses ( 3% to +28% ) , and when applying Ingenuity Pathway Analysis ( IPA ) up-stream analysis to ∼580 genes that co-varied with gain in lean mass , we identified rapamycin ( mTOR ) signaling associating with growth ( P = 1 . 4×10−30 ) . Paradoxically , those displaying most hypertrophy exhibited an inhibited mTOR activation signature , including the striking down-regulation of 70 rRNAs . Differential analysis found networks mimicking developmental processes ( activated all-trans-retinoic acid ( ATRA , Z-score = 4 . 5; P = 6×10−13 ) and inhibited aryl-hydrocarbon receptor signaling ( AhR , Z-score = −2 . 3; P = 3×10−7 ) ) with RET . Intriguingly , as ATRA and AhR gene-sets were also a feature of endurance exercise training ( EET ) , they appear to represent “generic” physical activity responsive gene-networks . For age , we found that differential gene-expression methods do not produce consistent molecular differences between young versus old individuals . Instead , utilizing two independent cohorts ( n = 45 and n = 52 ) , with a continuum of subject ages ( 18–78 y ) , the first reproducible set of age-related transcripts in human muscle was identified . This analysis identified ∼500 genes highly enriched in post-transcriptional processes ( P = 1×10−6 ) and with negligible links to the aforementioned generic exercise regulated gene-sets and some overlap with ribosomal genes . The RNA signatures from multiple compounds all targeting serotonin , DNA topoisomerase antagonism , and RXR activation were significantly related to the muscle age-related genes . Finally , a number of specific chromosomal loci , including 1q12 and 13q21 , contributed by more than chance to the age-related gene list ( P = 0 . 01–0 . 005 ) , implying possible epigenetic events . We conclude that human muscle age-related molecular processes appear distinct from the processes regulated by those of physical activity .
Discovery of the biological determinants of muscle mass and functional molecular phenotypes has substantial bearing on the fields of human performance ( e . g . hypertrophy , strength or endurance adaptations [1] , [2] ) and human health ( countering muscle atrophy and deconditioning occurring in older age or with conditions such as cancer cachexia [3] , [4] , respiratory disease or medically enforced immobilization ( e . g . hospitalized bed-rest , cast-immobilization [5] , [6] ) ) . Resistance exercise ( RE ) training ( RET ) is an effective intervention to increase muscle mass in many , but not all people , and thus provides an excellent opportunity to study gene-network regulation during muscle hypertrophy and the proposed relationship to muscle aging . Many exogenous factors may influence RET-induced hypertrophy including manipulation of exercise volume [7] , intensity [8] and adequate macronutrient availability [9] all of which interact with an individual's genotype to determine muscle growth . Establishing the molecular diagnostics that enable a personalized approach to tackle ageing has great appeal . To date , the molecular regulation of muscle hypertrophy has mostly focused on aspects of post-genomic signaling , with early work concluding that canonical insulin-like growth factor ( IGF-1 ) signals control muscle hypertrophy though a phosphatidyl-inositol-3 kinase/protein kinase B/mechanistic target of rapamycin ( PI3K/AKT/mTORc1 ) pathway [10] , abbreviated to mTOR . This protein complex can control cell growth through two mechanisms; firstly , mTORC1 regulates efficiency of translation through inducing phosphorylation of its substrates , ribosomal protein S6 kinase ( S6K1 ) and 4E binding protein 1 ( 4EBP1 ) and , secondly , mTORC1 increases translational capacity through regulating ribosomal RNA ( rRNA ) production within the nucleolus [11] . There are however conflicting data regarding the importance of mTOR regulation ( protein phosphorylation or target gene mRNA responses ) or its up-stream regulators , and acute anabolic or chronic growth responses to resistance exercise [12]–[17] reported from the same laboratories , indicating that important biological rather than methodological issues remain to be identified . More recent evidence indicates that the mechanisms regulating muscle hypertrophy go beyond the canonical IGF-1/PI3K/AKT/mTORc1 pathway . While circulating IGF-1 concentrations do not determine RET-induced hypertrophy in humans [18] , hypertrophy has now been shown to potentially occur through both PI3K-AKT [19] and mTOR [20] independent pathways , even in pre-clinical models . Perhaps the most convincing observation in favor of a more divergent regulation of muscle growth , is the fact that disparate exercise modes ( e . g . RET vs . endurance exercise training ( EET ) ) can produce similar protein signaling patterns in humans [21] . This suggests that the molecules , so far studied , are pleiotropic and in our opinion probably important for any type of tissue remodeling , regardless of the final physiological phenotype [22] . Therefore , a more innovative approach is needed to define links between molecules and ensuing in vivo physiological adaptations , than can be achieved with targeted western-based molecular profiling . Exercise training has also been postulated as a key tool to reverse the impact of ageing on human skeletal muscle phenotypes [23] , [24] . Yet , while some ‘genomic features’ of ageing have been reported [25] , we have noticed that the available global molecular profiles of human muscle [23] , [24] , [26] , [27] do not identify consistent molecular features . Furthermore , our recent work has highlighted that physiological adaptations to exercise , whether that be hypertrophy [28] or aerobic function [29] , are highly heterogeneous in humans , implying that exercise may not be able to “reverse” muscle ageing [23] for some people . For example , following >10-wk of supervised EET , ∼20% of subjects show no improvement in aerobic capacity while ∼30% demonstrate no improvement in insulin sensitivity [30] , [31] . Similarly , we reported muscle hypertrophy ranging from 0 . 8 to 6 . 0 kg [28] , while Raue et al reported changes in muscle cross-sectional area ( CSA ) ranging −1 . 2 to +10 . 4 cm2 [24] . Both of these RET studies reported that the outcome of supervised progressive RET did not relate to pre-existing differences in characteristics ( i . e . , gender , age , pre-existing muscle mass , physical activity levels or dietary habits ) indicating that there is not a simple explanation for the heterogeneity of the gains in lean mass [28] , [32] . In recent years , we have focused on using the heterogeneous responses to exercise training and OMIC techniques to uncover molecular networks regulated by EET [33][29] or generate molecular predictors of trainability [34] , directly in humans . The aim of the present work is to produce the first reproducible molecular signature of human muscle age , and examine how such a profile relates to new and established exercise adaptation gene networks . We generated new gene-chip profiles from muscle samples derived from two independent clinical cohorts , with a continuous range of ages ( 18–79 y ) and which originate from distinct environments ( UK and USA ) and which were independently processed in the laboratory . We also generated a new data set of paired global RNA-responses to a supervised 20-wk RET program ( N = 44 ) , as well as utilizing various sets of published acute-RET and chronic-EET gene-chips ( total N = 200 ) data sets . Finally , Ingenuity's new IPA up-stream analysis tool [35] was used to identify key features , within these novel age and exercise signatures , to provide independent and robust molecular insight into the heterogeneous nature of muscle hypertrophy , and human muscle age .
Of greater physiological importance was our effort to identify genes which link to the magnitude of muscle hypertrophy in humans . Quantitative SAM analysis [3] , [36] identified 642 probe-sets and interestingly the majority of genes identified were negatively correlated with gains in lean mass ( Dataset S2 ) . The probe-set list was imported into IPA ( filtered at a 5% FDR ) and 384 genes could be mapped to the IPA database for pathway analysis . The enrichment score generated for the EIF2 canonical pathway enrichment was extremely significant ( P-value = 1×10−64 ) and the combination of these observations indicates that the gene-list was both predictable ( valid with extensive literature ) and statistically very robust ( Figure S2 ) . We identified a number of regulators that could be responsible for regulating the transcriptional signature that correlated to gain in lean mass ( Dataset S2 ) . Figure 2A presents the most striking finding , an active rapamycin signature , equating to inhibition of mTOR signaling [37] . This signature was comprised of genes that almost entirely negatively correlated with lean-mass gain ( Z-score = 2 . 8 for directional consistency; P-value for transcript overlap p = 1 . 4×10−30 ) . In short , subjects that demonstrated the largest gains in lean tissue mass following 20-wks of RET have suppressed mTOR signaling over the training period , a novel and controversial observation ( all raw data was manually checked for consistency of direction ) . A second major transcriptional regulator was MYC , and the gene-list driving its inclusion ( Z-score −5 . 8 for directional consistency; p-value for transcript overlap p = 4×10−43 ) overlapped with the rapamycin network , equating to the inhibition of MYC activity . MYC can be an upstream activator of mTOR signaling in cell culture systems [38] . Thus these two robust observations are consistent , and notably the signature evidence is based on entirely independent data . Close examination of the rapamycin associated gene network ( Figure 2A ) reveals ∼30 ribosomal RNA genes ( a total of 70 genes were in the lean mass gain associated list but not all featured in the data-base of rapamycin regulated genes ) . The remaining genes ( Figure 2A ) belonged to metabolic processes and other facets of protein metabolism or signaling . To more easily appreciate the characteristics of those subjects that were found to have the greatest increase in lean tissue mass combined with a reduction in ribosomal gene abundance , we plotted the quartile response in lean mass ( Figure 2B ) . Baseline lean mass could not explain our observation and in fact the four groups had the same mean age , physiological characteristics , while the highest and lowest quartiles for lean mass gain had exactly the same proportion of males and females ( Table 1 ) . Figure 2C specifically illustrates a subset of the ribosomal RNA genes , however all other rRNA genes were consistent with this plot . There was an almost universal pattern of down-regulation observed in high responders ( shown in purple ) while subjects that were unable to increase muscle mass ( shown in blue ) substantially up-regulated these genes , as would be expected from the qSAM statistical analysis ( FDR<5% ) in Figure 1A . Thus , our analysis strategy enabled the discovery of an entirely novel in vivo feature of the mTOR growth pathway , while standard differential RNA expression analysis ( pre vs . post sample ) could not . We also plotted the relationship between physiological characteristics , protein-phosphorylation during acute net anabolic situations ( resistance exercise coupled with feeding ) with lean mass gain in these subjects using principal component analysis ( PCA ) . Selected variables were scaled and principal components 1 and 2 are presented . In Figure 3A , it is abundantly clear that none of the metabolic or physiological characteristics shared variance with the main component capturing lean mass gain variation following RET . Likewise , while protein kinase abundance or protein kinase activation status varied in a manner consistent with the literature , none of these acute net anabolic responses were correlated to gains in RET induced lean mass or shared variance with gains in lean mass when studied prior to 20-wk RET ( Figure 3B ) . In short , only the unbiased transcriptomics method was able to identify a biological profile distinguishing between high and low responders for lean mass gain . Identification of the determinants of skeletal muscle mass has obvious implications for treating age-related sarcopenia . There is no longitudinal molecular analysis of ageing muscle in humans . However using cross-sectional gene-chip data-sets , attempts have been made to identify age-related gene expression changes . For example , Melov et al . , reported that differences in gene expression between a cohort of young and old subjects can be removed through a RET program [23] . RET removed some aspect of the ‘inactivity’ related component of the difference between young and old subjects but when we contrasted their age-signature with other publically available muscle age datasets [23] , [24] , [26] , [27] no overlap was apparent . To investigate this issue further , we utilised the Melov et al . , data and the data from Trappe and our lab . We used SAM analysis to compare young with old subjects in each study . For the Trappe study we had 13 young ( 20–30 y ) and 11 old subjects ( >80 y ) . We used baseline samples from DRET , 10 young ( 20–29 y ) versus 16 old subjects ( 64–75 y ) , and finally we re-analysed the Melov et al . , data [23] which had 26 younger ( 18–28 y ) and 25 older subjects ( 65–85 y ) . As can be clearly observed in Figure 4 there was no overlap between the three studies indicating that a reproducible set of ‘ageing’ genes cannot be generated with this statistical or experimental approach . Our re-analysis of the GO analysis ( using DAVID ) of earlier studies [26] , [39] , using the appropriate back-ground files [40] , also confirmed that mitochondrial RNA changes in ageing cannot be claimed as being a reproducible hallmark of ageing , despite the presumed association with inactivity . Re-analysis of the Melov et al . , data did identify a mitochondrial gene expression signature ( less significant than the original analysis due to comparison with a more appropriate ontological background ) but in that study the older subjects were substantially de-trained and aerobic function was not presented . This gene-set is also known to reflect physical activity [29] and inactivity in humans [41] and thus it shouldn't be attributed to age per se anyway . To consolidate the conclusion that there was no common finding across these three studies , we entered the individual gene-lists in a gene ontology analysis to evaluate if some common pathways were enriched in each list , even if the member genes differed . Only 1 ontological group was common to 2 from the 3-gene lists and it related to mitochondrial processes indicating that even when the older subjects have a lower physical capacity a decline in mitochondrial genes is not always a prominent feature of age-related changes . Therefore , an alternative approach to identify age-related gene expression profiles in human muscle was required . To achieve this we utilised QSAM , which we have previously applied and validated to some extent for human studies [3] . We applied QSAM to generate a list of muscle transcript levels , which correlated with subject age ( 20–75 yr ) , with correction for multiple testing . This allowed us to identify genes that either negatively or positively associated with the subject age . This has not been attempted before because previous studies did not have a sufficiently wide and continuous range of ages to generate such data [23] , [24] , [26] , [39] . However , such an analysis would be of limited value if some of the observations could not be independently reproduced , using a distinct set of clinical samples . To this end , we generated 52 new U133+2 profiles ( 17–63 yr ) from muscle biopsy samples from the HERITAGE Family Study [42] and were able to identify a set of 580 genes or transcripts ( Dataset 3 ) which were correlated with age in both clinical studies and which was enriched in post-transcriptional and chronic disease traits ( Figure S3 ) but not mitochondrial related gene-sets . We found in IPA that the age-related dataset was consistent with the activation of the PGR ( z-score = 2 . 6 and P-value = 0 . 001 ) and RXR ( z-score = 2 . 0 and p-value = 0 . 0001 ) proteins and 5-fluorouracil agonism ( Z-score = 2 . 2 and P-value = 0 . 0005 , Figure 5A ) . Each mediator orchestrated a set of either positive ( yellow ) or negatively ( green ) age-correlated genes such that both overlap and direction were similar to the literature-constructed networks . Critically , these networks were not significantly related to the lean-mass associated gene-list ( Figure 5B ) or differentially regulated by either acute RE ( Figure S4A ) or chronic endurance exercise ( Figure S4B ) . Thus it is unlikely that these new age-related observations reflect simple confounding factors , such as exercise training or being physically active . There was also inhibition of certain protein mediators with age ( Figure 6A ) including c-MYC ( z-score = −2 . 8 and p-value<0 . 0001 ) and CLDN7 ( z-score = −2 . 6 and p-value = 0 . 05 ) . Again , no clear relationship with acute exercise or endurance training was apparent ( Figure S5 ) , while a closer association with genes related to gains in lean mass was noted ( Figure 6B ) with some key exceptions . Notably inhibition of MYC was predicted in both the lean-mass and age-related gene lists ( with gene-correlations in the same direction ) which we would not expect if muscle ageing was simply the ‘opposite’ of muscle growth or lack of response to exercise training . Furthermore , large differences in gene expression still existed when comparing the age groups and the pre and post-training samples in the Trappe dataset ( data not shown ) . The age-related expression signature was also related to RNA signatures in the Broad Connectivity database , including multiple serotonin antagonists and appears opposite to DNA topoisomerase inhibition ( Dataset S4 ) . Finally , we examined whether the age-related genes were over represented at genomic loci using Positional enrichment analysis [43] . Both Chromosome 1 ( q12 ) and 13 ( q21 . 33 ) had significant hits and the genes associated with those locations can be found in Figure 7A and 7B and the remaining analysis in Dataset S5 .
Muscle hypertrophy is the most recognized adaptation to RET . However , there are numerous other adaptations that occur , to support the biochemical , physical and metabolic requirements of a growing muscle . For example , hypertrophy is associated with activation of muscle satellite cells to support growth [45] while RET stimulates angiogenesis proportional to muscle fiber growth ( rather than increasing capillary-to-muscle fiber area as EET [46] ) . Furthermore , along with accretion of contractile proteins , the extracellular matrix ( ECM ) undergoes substantial remodeling after RET [47] . Clearly then , successful hypertrophy is the summation of complex intra/extra-muscular cross talk to co-ordinate hypertrophy . As such , we believe that teasing out the molecular networks regulating adaptation in vivo requires charting the relationship between the modulation of molecular factors with that of the physiological outcome ( s ) [29] . Our initial analysis revealed that activation of a Tretionin ( all-trans retinoic acid ( ATRA ) ) and inactivation of aryl hydrocarbon receptor ( AhR ) are common molecular responses to training , irrespective of exercise mode i . e . , RET or EET . ATRA is the active form of vitamin A , which serves as a ligand for two families of widely expressed nuclear receptors; the retinoic acid receptors ( RAR ) that bind to ATRA and the retinoid X receptors ( RXR ) that bind to its stereoisomer , alitretionin ( 9-cis-RA ) . Although little work exists on ATRA signalling in skeletal muscle , Halevy et al . showed exogenous ATRA promoted myogenic cell differentiation [48] which is allied to the function of ATRA amongst various cell types [49] . Given the post-mitotic properties of myonuclei , this may point to a novel link between ATRA-like signalling and aspects of in vivo satellite cell activities induced by exercise training . In support of this idea , we also observed IGF-1 and IGF-2 up-regulation in both RET and EET , and this is thought to play a role in satellite cell activation and differentiation [50] , [51][52] . ATRA-like signalling has also been shown to modulate endothelial cell maturation and angiogenesis in tube formation assays [53] suggesting a role for activation of this network in exercise-induced angiogenesis , while angiogenic factors are also associated with satellite cell activation and differentiation [54] . Indeed , we identified up-regulation of homeobox ( HOX ) genes ( e . g . HOXB3/7 ) , which have roles in vascular remodelling and angiogenesis [55] . Thus this collection of genes is likely contributing to vascular remodelling-induced both by RET [56] and EET [57] . In addition , the turnover of ECM components is activated when smooth muscle cells are exposed to exogenous ATRA , thereby suggesting this pathway is involved in ECM remodelling [58] . Indeed , the ( activated ) ATRA gene list was highly enriched in collagen genes . Finally , while one cannot rule out that the degree of modulation of ECM ( which differed between RET and EET ) may influence aerobic adaptation [29] , [31] it would appear to be a constituent feature of muscle growth and remodelling per se . The AhR is a ligand-activated transcription factor known to mediate the negative effects of environmental xenobiotic contaminants such as dioxin ( i . e . , TCDD; 2 , 3 , 7 , 8-tetrachlorodibenzo-[p]-dioxin ) . This receptor belongs to the basic-helix-loop-helix ( bHLH ) /PAS ( Period [Per]-AhR nuclear translocator [Arnt]-Single minded [Sim] ) family of heterodimeric transcriptional regulators [59] . There have been a number of studies in which physiological clues have been gathered as to the function of AhR . For example , AhR has shown tumour suppressor effects i . e . , when receptor levels are down-regulated by siRNA [60] the G1/S transition of the cell cycle and cell proliferation is promoted , suggesting a growth inhibitory role of the receptor . There also exists links between AhR and angiogenesis . Vascular endothelial growth factor ( VEGF ) , a major growth factor that regulates angiogenesis is transcriptionally regulated by hypoxia-inducible factor-1 alpha ( HIF-1α ) in response to tissue hypoxia and muscle contraction [61] . Hypoxia stabilizes HIF-1α , which forms heterodimers with HIF-1β , also known as the AhR nuclear translocator ( ARNT ) . ARNT can heterodimerize with AhR , minimizing the ability of ARNT to interact with stabilized HIF-1 to induce VEGF production . Finally , matrix remodelling is negatively affected during AhR activation . For example , AhR activation blocks regenerative processes during zebra fish caudal fin regeneration while impairing expression of genes involved in the structure and remodelling of the ECM [62] . Thus , inactivation of AhR-signalling may facilitate ECM and vascular remodelling occurring in response to both RET and EET ( Figure 1 ) . The notion that both ATRA and AhR pathways are potent regulators of cell growth , differentiation , ECM remodelling , vascularization , organogenesis and embryogenesis [63] underlines their key roles in tissue development and homeostasis . Moreover , since adaptation to RET and EET both involve aspects of cellular remodelling , muscle satellite cell activities [64] , ECM and vascular remodelling , we suggest that ATRA and AhR molecular programs are playing a previously undefined but central role in regulating “generic” features of exercise adaptation [22] . Finally , since ATRA and AhR gene networks that were regulated during long-term exercise training ( Figure 1 ) , were not reflective of those modulated in the hours after a single bout of exercise [24] , [33] , this casts doubt over ascribing formative purpose of acute exercise gene networks , which more likely represent stress pathways instigated by unfamiliar activities or simply the acute energy crisis in exercised muscle ( in agreement with the lack of a striking ontology profile ) . This may explain why acute mRNA changes do not overlap with the chronic exercise patterns or , in our hands , relate to the networks that associate with the degree of gain in lean mass ( see below ) . Following the identification of what might be called generic ‘adaptability’ [22] molecular networks , we were interested to see if we could identify any molecular networks that were regulated in proportion to the degree of muscle hypertrophy in individual subjects . The justification for this approach was based on the marked heterogeneity in capacity for muscle growth in humans , with gains ranging from 0% to 22% [24] , [28] , [65] . In the DRET study , we found a similar range of changes in muscle size ( −3% to +28% ) and the analyses of the gain-related gene networks yielded striking results . We first discovered that there existed a correlation between capacity for human muscle lean mass gains and activity of both c-MYC and mTORc1 sensitive genes , including a large group of ribosomal RNA ( rRNA ) genes ( ∼70/560 total rRNAs ) . We further demonstrated that the nature of the association was not as one would be expecting from pre-clinical research , but rather there was a reduction in rRNA gene expression when greater muscle hypertrophy was observed ( Figure 2B ) . We speculate that human high-responders to hypertrophy potentially show superior efficiency in protein synthesis ( i . e . , protein yielded per RNA ) and/or a reduction in proteolytic responses to RET . Nonetheless , it should also be noted that while rRNA expression negatively correlated to gain in lean mass , the highest responders tended to have higher levels of rRNA pre-training for individual rRNAs and hence the absolute abundance was similar post-training . This pattern of response clearly demonstrates that ‘more’ mTOR activation ( e . g . more rRNA production ) in humans is neither a hallmark nor a necessity for gains in lean mass in vivo . Our observations in humans also contrast with the molecular responses observed during acute synergist ablation induced hypertrophy [66] , raising further doubts over the relevance of such pre-clinical models to inform about physiological muscle growth in humans . Regardless , our data demonstrate that high-responders for muscle hypertrophy evoke an “anti-growth” transcriptional response during a period of successful muscle growth with more studies being required in high or low lean-mass responders to investigate this phenomenon unambiguously . While the unbiased transcriptional profiling yielded novel insights into human muscle growth responses , we also attempted to link training responsiveness to subject baseline physiology and the acute response of phospho-protein ( AKT-mTORc1 ) signaling in response to an anabolic stimulus ( RET combined with optimal nutrition [67] ) . We used correlation analysis to examine the relationship between key factors that have been speculated to be important for human muscle growth ( e . g . body composition , fiber-type , metabolic and signaling molecule status ) . We found no significant shared variance between any of these variables and gains in lean mass , as presented in Figure 3 using PCA . We utilized PCA for visualizing the integration of physiological and molecular data because it enables an over-view in a single plot of how physiological and molecular aspects may vary within the major ‘units’ of variance of a particular dataset rather than plotting multiple individual scatter plots . By doing this , it becomes easier to visualize when paired relationships are behaving as expected , such as total protein and phospho-protein positions within each principal component . Unusual patterns could be used to identify problem areas such as when detection methods ( e . g . antibodies ) are poorly functioning . As can be seen in Figure 3A , while principal component 1 was dominated by the variation in lean mass gains , basal lean mass , fat mass or fasting glucose did not vary . These data are in agreement with previous studies in which baseline physiological status was unable to select out high-responder status to hypertrophy [28] . We integrated the acute changes in phospho-protein signaling under the conditions of RET plus nutrition as an index of growth signaling potential in each subject . This seemed a valid approach as muscle hypertrophy is the product of nutrition and exercise-induced muscle protein synthesis [68] . While pre-training acute “anabolic” signals did not share variance with gains in muscle lean mass ( Figure 3A ) , some weak relationships appeared between acute “anabolic” signaling elements following exposure to RET ( Figure 3B ) and gains in lean mass . Collectively , this underlines that while acute changes in phosphorylation may associate with those of acute remodeling processes ( i . e . , muscle protein synthesis responses ) they are a poor indicator of future leans mass gains . Previously we have found a degree of dissociation between AKT-mTOR signals [69] , [70] and muscle protein synthesis , while others have shown that acute synthesis does not , but some signaling molecules do , relate to future gains in lean mass [71] . We contend that using these individual signals as acute proxies for RET muscle growth is not going to be the most sensitive strategy . There is a long established relationship between canonical pathways related to muscle growth and mechanisms that are associated with extended life-span in model organisms [72] . For example , inhibitors of mTORc1 and PI3K ( a gene also known as “age-1” ) activity can extend life-span in Caenorhabditis elegans , drosophila and mice [72] . As the number of people living beyond their eighth decade rises , it is expected that skeletal muscle atrophy and dysfunction ( sarcopenia and dynapenia respectively ) will become an increasing public health challenge [73] . Activation of mTORc1 and muscle IGF-1 signalling is associated with muscle cell growth experimentally while chronic inhibition of mTOR has been predicted to induce muscle frailty in humans [14] , [74] . Likewise , reduced S6K1 ( ribosomal protein S6 kinase ) activity of the multifaceted regulator of cell growth would be thought to impair the retention of muscle mass in humans [75] . Thus there is a clear molecular basis to believe that processes under-pinning ageing , longevity , sarcopenia and muscle growth will be strongly inter-connected . We found a list of ∼500 genes which track with age in human muscle across two independent cohorts , when our analyses utilised a continuum of ages . This contrasts with the irreproducibility of previous RNA versus muscle age datasets . Using this data we then evaluated which age-related gene networks link to muscle growth signalling or a variety of exercise scenarios [24] , [29] , [56] , [76] . The motivation for this is that these factors cannot be independently controlled for in human studies and thus post hoc considerations are essential . The first clear observation was that the reproducible age-related gene-list and the lean-mass gain related gene-list both had inhibition of MYC as a key transcriptional feature immediately indicating that age and muscle growth are not exact opposites . Further , the key upstream regulators of the age-related gene list ( e . g . PGR , RXR , Claudin7 ) contained a set of genes which were largely unrelated to the RET and EET transcriptomes , or the acute exercise responses [24] , [76] for that matter . Claudin-7 activity was inhibited with age ( Figure 6 ) and appears to relate to developmental differentiation and is strongly regulated by androgen signaling and HGF in vitro [77] . According to the IPA publication database the RXR ligand , CD437 , induces a transcriptional signature that is consistent with age-related gene correlations , involving a network that is for 2/3rd negatively correlated with increasing age and for 1/3rd positively associated with age ( i . e . with age KLF5 and IRS2 expression increases ) . Note that the ATRA signature , characteristic of increased physical activity in our hands , should operate through the RAR pathway . However , there is some overlap in the down-stream gene activation and thus both age-related changes and exercise activate some common features related to Vitamin A biology . In short , it is abundantly clear that the age-related changes in gene expression are not simply the ‘opposite’ [23] of the profiles seen with exercise or exercise training . Positional enrichment analysis [43] was used to map the ∼500 age-related genes to chromosomal localisation , as attempts to link DNA variants with ageing have so far been only partly successful suggesting that an alternative approach may provide useful insights . We found several loci that yielded a significant enrichment score with ∼1q12 ( e . g . PDE4DIP ) and ∼13q21 ( e . g . LMO7 ( FBXO20 ) ) yielding the most significant scores ( Figure 7 ) . PDE4DIP is a binding partner of phosphodiesterase 4D ( PDE4D ) which partners with Rheb to be a cAMP-specific negative regulator of mTORc1 [78] . When PDE4D binds Rheb it inhibits the ability of Rheb to activate mTORC1 and hence it is plausible that PDE4DIP ( also called myomegalin ) impacts on this relationship , as an increased interaction between Rheb and mTOR should promote growth . On chromosome 13 , another growth and differentiation gene , Lmo7 , was identified . Lmo7 impacts on myoblast differentiation , being required to induce Pax3 and MyoD expression [79] . Lmo7 is positively associated with age in muscle , while we have previously identified PAX3 as an up-stream transcriptional regulator of the EET induced transcriptome [29] , highlighting further why the age-related transcriptome is not simply the opposite of exercise training . i . e . , muscle ageing is not simply inactivity and thus is unlikely to be reversed by only activity ( this does not at all contest that muscle function can be substantially retained by physical training in many but not all subjects ) . Over the past 2 decades , therapeutic advances for complex chronic diseases have failed to generate all the progress predicted by the emergence of genetic technologies [80] . Part of the reason for this is that the investment in forward genetic pre-clinical models [81] has not yielded the expected validation of drug targets and it is now widely accepted that the clinical approach to chronic-disease management will have to reflect on numerous interactions between environmental and molecular factors [82] . Thus , an alternative approach , whereby identification of disease networks directly observed in clinical populations may have merit and lead to a more rapid translation of basic science [3] , [4] , [34] , [40] , [83] . Validation of such observations is dependent on access to sets of independent clinical data sufficiently large to have robust statistical power and diverse enough to be able to generalize the conclusions . It is safe to say that our approach is not at all universally favoured , as there is still a great reliance on so-called ‘validation’ studies involving forward and reverse genetic strategies in mice and cells . We suspect that such validation studies will , in the end prove to be context dependent and no easier to interpret than our in vivo molecular studies . The singular advantage of our approach is that our data is generated under the precise conditions that would ultimately require therapeutic intervention . However , in the present analysis we focus on components of dynamic function , namely muscle mass , as the study of muscle performance will require determination of a wider range of muscle functional parameters ( power , torque , velocity ) and larger clinical studies , studies which do not exist at present . We also appreciate that our current observations would benefit greatly from follow-on genetic association studies in humans and by pharmacological or nutritional intervention studies . For example we have shown that a simple relationship between more mTOR activation and muscle growth does not exist in vivo in humans , albeit we are relying in part on the pharmacological activity of rapamycin to support this observation ( off target effects may be present ) and do not yet have kinetic data to understand the dynamic nature of this new relationship . We also failed to find a link between inter-subject variation in acute phospho-protein anabolic signalling and gains in lean mass . This may reflect the choice of time-point that we profiled the protein responses under . However , it must also be recognised that quantification of protein abundance changes suffers from numerous complications , including compression or exaggeration of dynamic range and challenges with specificity . Time will tell if our systems-biology translational medicine approach [84] exceeds the performance of traditional approaches taken to yield new therapeutic advances for human health .
Subjects were recruited from an age range of 18 to 75 y . Before beginning the study all subjects were screened using a medical questionnaire , physical examination and resting ECG with exclusions for overt muscle wasting ( >2 SD below age norms ) [85] , metabolic , respiratory/cardiovascular disorders or other major contraindications to a healthy status . All subjects had normal blood chemistry and were normotensive ( BP<140/90 ) . All subjects performed routine activities of daily living and recreation but did not participate in moderate to high intensity aerobic exercise and none had participated in RET in the last 24 months . Body composition was measured at screening and following RET by dual energy X-ray absorptiometry ( DEXA ) ( Lunar Prodigy II , GE Medical Systems ) . Subject positioning on the DEXA bed was optimized to allow the region of interest ( ROI ) body compartments to be analyzed separately . The upper leg ROI was selected as the area inferior to the lowest visible point of the coccyx to the mid-point of the patella . All subjects gave their written , informed consent to participate after all procedures and risks were explained . This study was approved by the University of Nottingham Medical School Ethics Committee and complied with the Declaration of Helsinki . The clinical data from these subjects were first reported in 2012 [56] . For the purposes of this article a total of 45 from the original 51 subjects were utilised as this represents the total number of gene-chip profiles that passed the appropriate quality control processes ( N = 89 U133+2 Affymetrix chips ) . The 20-wk fully supervised RET programme was designed to achieve skeletal muscle hypertrophy . Subjects trained three times per week , with each session lasting approximately 60 min . During four weeks of induction training ( to ensure adoption and adherence to correct technique ) intensity was increased from 40% to 60% 1-RM . For the remaining 16-wk of training intensity was set at 70% 1-RM with multiple sets of 12 repetitions , with two min rest between sets . 1-repetition maximum ( 1-RM ) assessments were made every four weeks to ensure that the intensity of loading was constant . Subjects were excluded from the study for non-compliance , defined as: non-attendance for >6 consecutive sessions , less than 75% attendance , or failure to complete the set exercise regime on >15% attendance . Muscle biopsies ( ∼150 mg ) were taken under fasted-non-exercised ( “basal” ) and optimal growth conditions ( acute exercise-fed conditions , 2 . 5 h after a single bout of exercise ) both before and after chronic-RET from the vastus lateralis muscle under local anaesthesia ( 2% lidocaine , with the use of a conchotome biopsy forceps , as previously described [56] ) . Blood was collected in pre-chilled tubes containing Lithium Heparin , plasma was separated by centrifugation and was then stored at −80°C until analyses . Plasma glucose concentration was measured on a clinical chemistry glucose analyser ( ILAB 300 Plus ) . For insulin , blood was collected in pre-chilled tubes containing EDTA , plasma was separated by centrifugation within 30 min of collection and was then stored at −80°C until final analyses . Plasma insulin concentration was determined using high sensitivity insulin ELISA systems ( Immunodiagnostic systems limited ) . Muscle biopsies ( ∼20 mg ) were homogenized in ice-cold extraction buffer ( 10 µL . mg−1 ) containing 50 mM Tris-HCl ( pH 7 . 4 ) , 0 . 1% Triton X-100 , 1 mM EDTA , 1 mM EGTA , 50 mM NaF , 0 . 5 mM activated sodium orthovanadate ( Sigma Aldrich , Poole , UK ) and a complete protease inhibitor cocktail tablet ( Roche , West Sussex , UK ) . Homogenates were centrifuged at 10 , 000 g for 10 min at 4°C . Bradford assays were used to determine supernatant protein concentrations after which samples were standardized to 1 mg . mL−1 in Laemmli loading buffer . Samples were heated at 95°C for 5 min before 15 µg of protein/lane was loaded on to Criterion XT Bis-Tris 12% SDS-PAGE gels ( Bio-Rad , Hemel Hempstead , UK ) for electrophoresis at 200 V for ∼60 min . Gels were equilibrated in transfer buffer ( 25 mM Tris , 192 mM glycine , 10% methanol ) for 30 min before proteins were electro-blotted on to 0 . 2 µm PVDF membranes ( Bio-Rad ) at 100 V for 30 min . After blocking with 5% low-fat milk in TBS-T ( Tris-Buffered Saline and 0 . 1% Tween-20; both Sigma-Aldrich , Poole , UK ) for 1 h , membranes were rotated overnight with primary antibody ( all AbCam , Cambridge , UK ) against our target proteins ( AKT , mTOR , p70S6K1 , 4EBP1 , eEF2 ) at a concentration of 1∶2000 at 4°C . Membranes were washed ( 3×5 min ) with TBS-T and incubated for 1 h at room temperature with HRP-conjugated anti-rabbit secondary antibody ( New England Biolabs , UK ) , before further washing ( 3×5 min ) with TBS-T and incubation for 5 min with ECL ( Immunstar; Bio-Rad ) . Blots were imaged and quantified by assessing peak density , after ensuring bands were within the linear range of detection using the Chemidoc XRS system ( Bio-Rad , Hemel Hempstead , UK ) . Protein phosphorylation was corrected for loading to actin loading control before the protein signals were subject to PCA to explore relationships between ‘anabolic signals’ , specifically in terms of muscle hypertrophy responsiveness . Total RNA was isolated from muscle biopsies taken before and after ( 72 h after the final training session ) RET by chloroform-phenol based extraction . In brief , paired tissue samples ( obtained before and after RET ) of ∼20 mg each were processed simultaneously in 1 mL TRIzol ( Invitrogen ) using a Mini-Beadbeater-8 ( Biospec Inc . ) for 15 sec on the “homogenize” setting . After 5 min of incubation at room temperature , 200 µL of chloroform ( Sigma-Aldrich ) was added and samples shaken vigorously by hand . Samples were briefly incubated on ice prior to centrifugation at 12 , 000 g for 15 min . The supernatant was removed and mixed with isopropanol ( Sigma-Aldrich ) and spun once more at 12 , 000 g for 10 min , after 10 min of incubation . After a single washing step with 75% EtOH RNA pellets were re-suspended in 40 µL DEPC-treated water ( Ambion ) and quantified using a NanoDrop Spectrophotometer ( NanoDrop Technologies ) . RNA purity was assessed using the A260/A280 , A260/A230 ratios and stored at −80°C . Samples were put through this process in pairs ( pre-post samples ) while the order of subject processing was carried out to distribute ‘non-responders’ equally . Reverse transcription of RNA was carried out using the Affymetrix 3′ IVT express kit . 100 ng of total RNA was reverse transcribed as per manufacturer's protocol , and quantified using a Nanodrop ND-1000 instrument . aRNA was fragmented and labeled as per manufacturers protocol and hybridized to Affymetrix U133+2 arrays ( Affymetrix , USA ) . Arrays were washed , stained and scanned following Affymetrix standard procedures , using an Affymetrix 3000 7G scanner and Affymetrix 450 wash station . A visual inspection of each array was carried out . Low-level processing of all arrays was undertaken using Bioconductor in R . The Affy package was used to carry out MAS5 based normalization and generate present , marginal and absent ( PA ) scores . NUSE plots were generated and combined with PCA , outlier samples were identified where both the NUSE plot and PCA was supportive of its exclusion ( ∼2% of arrays ) . For baseline correlation analysis , all samples that passed QC were utilized ( N = 45 ) . This procedure was applied to the data-set originating from the Trappe laboratory ( GSE28422 , [24] ) and outliers removed from the dataset that failed the QC process , leaving n = 96 for analysis . Pre-exercise training muscle biopsy samples from the HERITAGE family study ( N = 50 ) were also analyzed to yield a second independent data set with a continuous span of age-ranges ( see below ) . The Trappe and HERITAGE datasets therefore represent independent datasets which we utilized , where possible , to validate the pathway analysis of our study . Such confirmation benchmarks results using thousands of data-points and is more desirable that targeted real-time qPCR confirmation ( where the gene selection is biased and the sample size inappropriate to make statistical conclusions ) . Annotation of all CEL files used ‘hgu133plus2cdf_2 . 9 . 1 . tgz’ while annotation of probe-set lists was then updated using the Ingenuity Pathway Analysis database , as of August 2012 . Our first objective was to identify the gene-networks regulated in proportion to gains in upper leg muscle mass ( hypertrophy ) , the same location as our biopsy sample . Such analysis relies on the established principal that adaptation responses ( for the majority of phenotypes ) to exercise training in outbred populations is highly variable , typically reflecting genetic and epigenetic variation and in genomic variation . We utilized quantitative SAM analysis [3] , [36] to generate a list of genes which vary in a positive and negative manner with changes in DEXA assessed upper leg lean mass . This was applied to PA filtered data and the statistical parameter generated is a q-value ( false discovery rate ) . This provided for the first time a candidate list of gene-changes that may exhibit primary or secondary influence over muscle growth in humans . The gene-list was then subject to IPA based pathway analysis and in particular the Upstream Analysis tool in IPA was utilized . This analysis has similarities to the Molecular Connectivity Database [86] where pre-existing collections of RNA signatures are compared with our lean-mass related gene list , and significant overlaps identified . An overlap P-value is generated based on the degree of overlap between the gene-set within the IPA database ( which reflects the RNA molecules changed in response to a ‘mediator’ such as a transcription factor or a drug ) and our data set , adjusting for data set sizes using the Fischer's Exact Test . We accepted a stringent P-value of p<0 . 001 as being significant . A second parameter is the activation “z-score” where the directional change in RNA is compared between the IPA mediator data-set and our lean-mass gain gene list . The z-score informs on whether the drug/protein mediator is likely to be ‘active’ or ‘inhibited’ during gains in lean mass . Thus , if we discovered that an antagonist is ‘inhibited’ in our analysis , this indicates that the drug target is activated . However , in the present study the data-input refers to genes , which positively or negatively correlate with lean mass gains e . g . if we find a “Statin” signature was inhibited , it is interpreted that HMG-CoA reductase regulated genes are negatively correlated with lean mass gain . The two-step process presented above generates a focused gene-list with a high statistical rigor for true positive associations . This type of analysis also utilizes the full range of physiological response observed , however it assumes that expression of important genes will relate in a linear manner to lean mass gain and thus can not discover all appropriate associations . We then contextualize the statistical findings both in terms of subject characteristics and through comparison of the response of these significant networks with independent gene-array data ( e . g . [24] ) . At this stage we utilized descriptive statistics , plotting the significant network genes as simple expression values relative to the quartile distribution of lean-mass gains to allow for clear discussion of the results . As these plots are based on the z-scores and P-values as above , no further statistical analysis is presented . Following identification of our primary objectives we then carried out a classic differential expression analysis using SAM . Given that we have established that chronic differential expression patterns , following exercise training , are dependent on the presence of physiological adaptation we removed 6 subjects that demonstrated no gain in lean mass . This yielded a list of differentially expressed genes that could then be compared with the RET gene-list generated from the Trappe laboratory data and our published exercise studies [24] . Secondary analysis , where subject age or baseline lean-mass was related to baseline gene-expression was carried out using quantitative SAM analysis as described above [3] , [36] . This allowed us to present comparisons of the RET gene-list with other modes of exercise , such as endurance exercise training or disease [3] , [29] and age-related analysis [23] , [24] . PCA was utilized to visualize the association between selected physiological and protein expression parameters and training induced changes in muscle lean mass . PCA was implemented in R , using prcomp ( ) command , which calculated a singular value decomposition and plots the selected principal components using the plot command in R . All data was individually transformed to a median value within that data set so that all variables were within a consistent data range . In each case the majority ( ∼65% ) of the total variance was captured by the first two principal components . Finally , positional gene enrichment analysis ( PGE ) was used to identify whether the classification genes ( or the classifier network genes ) were significantly enriched within given chromosomal regions [43] . This analysis is based on the following rules: Rule 1: it contains at least two genes of interest , Rule 2: there is no smaller region containing the same genes of interest , Rule 3: there is no bigger region with more genes of interest and the same genes not of interest , Rule 4: there is no larger encompassing region with a higher percentage of genes of interest , Rule 5: there is no smaller encompassed region with a better P-value , Rule 6: it does not contain any region having less than expected genes of interest . The approach of PGE exhaustively evaluates the over-representation at all chromosomal resolution levels simultaneously .
|
A fundamental challenge for modern medicine is to generate new strategies to cope with the rising proportion of older people within society , as unaddressed it will make many health care systems financially unviable . Ageing impacts both quality of life and longevity through reduced musculoskeletal function . What is unknown in humans is whether the decline with age , referred to as “sarcopenia , ” represents a molecular ageing process or whether it is primarily driven by alterations in lifestyle , e . g . reduced physical activity and poor nutrition . Because the details of such interactions will be uniquely human , we aimed to produce the first reproducible global molecular profile of human muscle age , one that could be validated across independent clinical cohorts to ensure its general applicability . We combined this analysis with extensive data on the impact of exercise training on human muscle phenotype to then identify the processes predominately associated with age and not environment . We were able to identify unique gene pathways associated with human muscle growth and age and were able to conclude that human muscle age-related molecular processes appear distinct from the processes directly regulated by those of physical activity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"genome",
"expression",
"analysis",
"functional",
"genomics",
"integrative",
"physiology",
"anatomy",
"and",
"physiology",
"physiological",
"processes",
"genetics",
"and",
"genomics",
"musculoskeletal",
"system",
"geriatrics",
"gene",
"expression",
"biology",
"systems",
"biology",
"physiogenomics",
"physiology",
"genetics",
"genomics",
"gene",
"networks",
"genomic",
"medicine"
] |
2013
|
Molecular Networks of Human Muscle Adaptation to Exercise and Age
|
Onchocerca volvulus is the nematode pathogen responsible for human onchocerciasis also known as “River blindness” , a neglected tropical disease that affects up to 18 million people worldwide . Helminths Excretory Secretory Products ( ESPs ) constitute a rich repertoire of molecules that can be exploited for host-parasite relationship , diagnosis and vaccine studies . Here , we report , using a range of molecular techniques including PCR , western blot , recombinant DNA technology , ELISA , high performance thin-layer chromatography and mass spectrometry that the 28 KDa cysteine-rich protein ( Ov28CRP ) is a reliable component of the O . volvulus ESPs to address the biology of this parasite . We showed that ( 1 ) Ov28CRP is a putative ganglioside GM2 Activator Protein ( GM2AP ) conserved in nematode; ( 2 ) OvGM2AP gene is transcriptionally activated in all investigated stages of the parasitic life cycle , including larval and adult stages; ( 3 ) The full-length OvGM2AP was detected in in-vitro O . volvulus ESPs of adult and larval stages; ( 4 ) the mass expressed and purified recombinant OvGM2AP purified from insect cell culture medium was found to be glycosylated at asparagine 173 and lacked N-terminal signal peptide sequence; ( 5 ) the recombinant OvGM2AP discriminated serum samples of infected and uninfected individuals; ( 6 ) OvGM2AP competitively inhibits MUG degradation by recombinant β-hexosaminidase A but not MUGS , and could not hydrolyze the GM2 to GM3; ( 7 ) humoral immune responses to the recombinant OvGM2AP revealed a negative correlation with ivermectin treatment . Altogether , our findings suggest for the first time that OvGM2AP is an antigenic molecule whose biochemical and immunological features are important to gain more insight into our understanding of host-parasite relationship , as well as its function in parasite development at large .
Human onchocerciasis also known as River Blindness is a tropical disease caused by the parasitic nematode Onchocerca volvulus and is the world’s second leading cause of infectious blindness after trachoma . The infective larval stages ( L3 ) are transmitted through the bites of infective blackflies of the genus Simulium . They eventually give rise to adults that dwell in subcutaneous tissues where they can survive for about 15 years , with adult females hatching about 1600 microfilariae daily [1 , 2] . About 18 million people carry dermal O . volvulus microfilariae ( Mf ) worldwide with 99% of them living in Africa . Mfs are responsible for severe itching or dermatitis experienced by about 6 . 5 million people and blindness affecting 270 , 000 . Current estimates hold that about 187 million people are at risk of being infected and about 1 . 1 million disability-adjusted life-years ( DALYs ) were lost in 2015 due to onchocerciasis [3–5] . Onchocerciasis remains a public health problem in endemic communities despite advances made by ivermectin treatment in reducing the burden of the disease . In fact , due to limitation in the different segments of the past and ongoing control programs , including the risks associated with loasis co-endemicity [6] , ivermectin resistance reported in some endemic foci [7–9] and the logistic/financial burden involved in implementing the mass drug administration programs [10] , it is increasingly accepted that ivermectin treatment alone will not be capable of supporting disease elimination . Moxidectin , a recently FDA approved microfilaricidal milbemycin macrocyclic lactone , ushers a glimmer of hope as it is reported to be more efficient with a low affinity to p-glycoprotein transporters [11] and a 20–43 days lifespan [12–15] . However , the effect of repeated treatment as well as treatment of Loa loa coinfected individuals still needs to be assessed for evaluation of long-term effects of moxidectin-based elimination programs [16] . Furthermore , existing diagnostic tools , mainly the skin snip test and the antibody based Ov16 test [17] , suffer drawbacks hence requiring supplementary tests for post treatment surveillance [18] . The search for novel drugs , vaccines and diagnostic tools therefore remain imperative for efficient surveillance of disease elimination [19] . The characterization of novel putative targets in the Onchocerca genome will help to meet these concerns . Genomics , proteomics and bioinformatics studies of Onchocerca have been reported and have indeed provided insights into molecules that could be useful in the context of disease diagnosis , therapy or vaccine design [20–24] . However , most of these high throughput researches call for more studies to characterize these proteins on an individual basis . Parasite Excretory Secretory Products ( ESPs ) constitute the set of molecules produced by the parasite into the host environment . Helminth ESPs are known to perform a wide range of functions including modulation of host immune response , leading to immune evasion by the parasite , remodeling of host tissue giving rise to nodule formation and alteration of host tissue nutritional status amongst others [25] . Furthermore , the immunomodulatory properties of helminth ESPs and their potential use to treat allergy and autoimmune diseases [26 , 27] as well as metabolic syndrome have been reported [28] . The exploitation of ESPs of Onchocerca in the search for novel drugs , vaccines , diagnostics as well as mediators of immune response have also been described [29–33] . Given that the Onchocerca nodule is known to be vascularized [34 , 35] , such molecules could reach host circulation thereby supporting investigations towards the detection of O . volvulus ESPs in body fluids . The ganglioside GM2 Activator Protein ( GM2AP ) of O . volvulus is one of such ESPs . In other species , GM2AP functions in vivo as an essential cofactor of N-acetylhexosaminidase A in degrading the ganglioside GM2 in the lysosome [36 , 37] . GM2AP acts as a biological detergent , solubilizes , binds and transports different lipids [37 , 38] . The human GM2AP is best studied; its crystal structure has been established and some functional roles attributed to its domain [39 , 40] . A nematode GM2AP was also identified ( in Trichinella pseudospiralis ) and its unusual characteristics reported . Contrary to known canonical GM2AP function , this parasite orthologue does not facilitate degradation of GM2 ganglioside by N-acetylhexosaminidase A , although it does inhibit phospholipase D activity . This was correlated to the absence of a domain implicated in binding to hexosaminidase A [41] . In the current study , we identified and characterized the O . volvulus GM2AP herein denoted OvGM2AP . Our findings on its expression during the parasite life cycle and on the humoral responses to recombinant OvGM2AP expressed and purified from insect cells suggest that OvGM2AP is an antigenic molecule whose biochemical and immunological features are important to gain more insight into our understanding of host-parasite relationship , and ultimately , to elucidate the function of this new GM2AP .
Ethical approval for this study was obtained from the Cameroonian National Ethics Committee for Health Research ( No 2015/01/543/CE/CNERSH/SP ) and administrative authorization was obtained from the Cameroonian Ministry of Public Health . A written consent was obtained from all participants employed in the study . For minors , the consent forms were validated by parents or guardians . Participation was entirely voluntary and individuals were free to opt out at their discretion without fear of community leaders or health practitioners . Ethical concerns working with rabbit are addressed by the GenScript Institutional Animal Care and Use Committee ( IACUC ) , licence # SYXK ( SU ) 2008–0021 . The IACUC is accredited by the AAALAC ( approval date: 6/19/2009 ) . O . volvulus and O . ochengi parasite material were collected at different stages of the parasites . O . volvulus nodules were obtained from patients in the Kombone Health area of the South-Western Region of Cameroon by a trained medical doctor and O . volvulus worms were obtained from these nodules as previously described [42] . Briefly , average nodular masses were digested in 0 . 5 mg/ml collagenase for 9 h at 37 oC shaking at 90 rpm after which the male and female worms were cultured in incomplete RPMI ( Gibco , USA ) supplemented with 0 . 25 mg/ml gentamycin and L-glutamine . Bloodless skin snips were also collected from the left and right knee of heavily infected individuals ( microfilaria ( Mf ) load greater than 150 Mfs per mg of skin ) and placed in incomplete culture medium for 12 h for Mf to emerge . The worms were pipetted and washed thrice with 20% percoll . L2 and L3 larval stages were obtained from infected and/or infective blackflies as described elsewhere [43] . Briefly , infected Simulium flies were captured and grown in the laboratory for seven days after which the thorax and head regions were excised to obtain L2s and L3s respectively . Infected cattle skins containing O . ochengi nodules were obtained from the abattoir in Douala and immediately brought to the laboratory where O . ochengi worms were obtained from these nodules as previously described [44] . These infected cattle skins were selected following a search on available skins obtained after regular slaughtering and skinning procedures carried out at the abattoir . Only motile worms were used for the extraction of RNA , protein crude extracts and ESPs . Protein crude extracts were obtained by crushing worm samples in lysis buffer ( 150 mM NaCl , 1 . 0% Triton X-100 and 50 mM Tris pH 8 . 0 ) supplemented with 1 mM Phenylmethylsulfonyl fluoride ( PMSF ) . ESPs were obtained by culturing parasites in vitro using incomplete RPMI medium ( Gibco , USA ) . After 16 h of culture of parasites ( 4 adult females per ml of RPMI medium , 50 adult males and 800 L3s per 300 μl of RPMI medium ) in vitro , the medium was collected , quantified by Bradford assay and used for SDS-PAGE and western blot analyses . Following examination of participants involved in the study by trained medical personnel , blood was collected from patients living in the endemic region of Kombone Health Area of the Mbonge Health district , South Western Region , Cameroon . O . volvulus is known to be specific in this region at the exclusion of other filarial infections . These patients ( OVS ) were selected on the basis of an established presence of Mf in skin snip biopsies and/or presence of clinical manifestation of onchocerciasis . The clinical and demographic data of these patients are indicated in S1 Table . Sera were obtained from blood samples by employing an established protocol [45] , diluted 1:2 in glycerol and stored at -20 oC . Sera obtained from individuals residing in an onchocerciasis hypo-endemic region ( HES ) of Huye , Rwanda , and from European subjects ( ESC ) , were used as controls . Sera were also collected from individuals in Bandjoun ( Bandjoun , Cameroon ) , a region which has been on constant ivermectin administration for over two decades . Sera were collected from individuals whose infection status had previously been well characterized [46] . Loa loa serum ( LLS ) samples were collected from the loiasis patients in the endemic regions of the Mvila division in the rain forest of the southern region of Cameroon in areas with high loiasis endemicity but <20% prevalence of onchocerciasis and with no ongoing CDTI programs . These subjects had been previously recruited for a study on the effects of albendazole on L . loa microfilariae [47] . These samples were obtained from the Centre for Research on Filariasis and other tropical diseases ( CRFILMT ) , Yaounde , Cameroon . In summary , all the different sera types used have been extensively characterized in a related study [23] . Serum samples for subjects infected with other nematode infections including Brugia malayi , Wuchereria bancrofti , Ascaris lumbricoides and Mansonella perstans were obtained from the filarial repository , courtesy of the laboratory of Dr . Steven Williams . The amino acid sequence analysis of OvGM2AP was done using Protparam [48] and the sequence alignment of OvGM2AP and homologous sequences from selected nematode species was done using PROMALS3D [49] . Structural prediction analysis was done using the Phyre2 online tool [50] . The phylogenetic tree of OvGM2AP sequences was generated by neighbor joining using PHYLIP 3 . 695 . The full-length OvGM2AP cDNA ( WormBase ID: OVOC1952 ) fused to a C-terminal 8x-His tag was cloned into pAC8_MF , a modified version of the pAC8_H transfer vector [51] . The resulting plasmid ( pAC8_MF_OvGM2AP ) was co-transfected in insect cells with AcMNPV viral DNA ( Bac10:KO1629 , Δv-cath/chiA-LoxP:DsRed ) linearized by Bsu36I to generate the recombinant baculovirus which was subsequently amplified and used to infect large scale cultures for protein production following standardized procedures [52] . Briefly , 1 μg of pAC8_MF_OvGM2AP mixed with 2 μg of linearized viral DNA in 750 μl of the buffer ( 25 mM HEPES , 140 mM NaCl , 125 mM CaCl2 , pH 7 . 1 ) was added dropwise to 750 μl of Grace’s medium supplemented with 10% of fetal bovine serum ( FBS ) to obtain a calcium-phosphate precipitate which was allowed to form during 15 min . The precipitate was layered onto 2 x 106 Sf9 cells ( Novagen ) grown in Grace’s Medium ( G8142-SIGMA ) previously seeded in a 25 cm2 flask . After 4 h incubation at 27°C , the medium was changed and cells further incubated for 5 days at 27°C . The resulting culture supernatant constitutes the initial virus stock ( V0 ) , which was used to obtain the first ( V1 ) and second ( V2 ) amplification . For large scale OvGM2AP production , Sf21 insect cells ( IGBMC , Strasbourg ) cultivated in the serum free Sf-900 II ( Gibco ) medium were infected at a density of 1 . 0x106 cells/ml and MOI of 5 with the V2 stock . After 3 days of infection , cells were pelleted by centrifugation ( 1000 x g for 10 min ) and the culture medium was incubated with Ni Sepharose 6Fast Flow ( GE Healthcare ) at 4 oC for 2 h using 1 mL of resin for 1 L of culture . Following three washes using a 35 mM Imidazole solution in 10% glycerol , 20 mM HEPES pH 7 . 0 , 300 mM NaCl , 5 mM 2-mercaptoethanol ( Buffer A ) and , bound proteins were eluted with the same buffer in the presence of 500 mM Imidazole . Eluted proteins were dialyzed against Buffer A and snap frozen for storage at -80°C . Purification of OvGM2AP_8His contained in the clarified lysate was also tested . Cell pellets were re-suspended in a buffer containing 5 mM Imidazole , 10% glycerol , 20 mM HEPES pH 7 . 0 , 300 mM NaCl , 5 mM 2-mercaptoethanol and cells were lysed by sonication at 4°C , 40% Amplitude for 10 sec . The lysate was centrifuged at 20000 g for 30 min and incubated with Ni Sepharose 6Fast Flow ( GE Healthcare ) . Western blot analysis of the poly-histidine tag from recombinant OvGM2AP_8His was performed using the mouse anti-His antibody 1D11 , ( IGBMC antibody facility , Strasbourg , France ) for primary detection , the donkey F ( ab' ) 2 anti-mouse IgG ( H+L ) conjugated with Horse Raddish Peroxidase ( HRP ) as secondary antibody ( Interchim SA ) and the Super Signal West Pico Chemiluminescent substrate for detection of HRP ( Thermo Fisher scientific ) . Chemiluminescence was detected using the Amersham imager 600 QC ( GE Healthcare , Sweden ) . For glycosylation analysis , purified OvGM2AP was treated with Peptide:N-glycosidase F ( PNGase F ) ( Promega , USA ) using a protein/glycosidase ratio of 19/1 ( w/w ) for 2 h at 37°C . Deglycosylated and undigested OvGM2AP proteins were resolved by SDS PAGE followed by either Coomassie staining or western blot to analyze a band displacement . For mass spectrometry analysis , the proteins were digested by trypsin ( after reduction with 10 mM DTT for 1 h at 57°C and alkylation for 45 min in the dark with 55 mM iodoacetamide ) . Tryptic digests were analyzed using an Ultimate 3000 nano-RSLC ( Thermo Scientific , San Jose , California ) coupled with a linear trap Quadrupole ( LTQ ) -Orbitrap ELITE mass spectrometer via a nano-electrospray ionization source ( Thermo Scientific ) . Peptide mixtures were loaded on a C18 Acclaim PepMap100 trap-column ( 75 μm IDx 2cm , 3μm , 100 Å , Thermo Fisher Scientific ) equilibrated with 3% acetonitrile and 0 . 1% formic acid in H2O , and then separated on a C18 Accucore nano-column ( 75 mm internal diameter ( ID ) x50 cm , 2 . 6 mm , 150 Å , Thermo Fisher Scientific ) with a 120 min linear gradient from 3 to 80% acetonitrile in the same buffer . Peptides were analyzed by Top 20-CID ( collision induced dissociation ) data-dependent MS . Spectra were processed with Proteome Discover 2 . 1 against a protein sequence database for Sf21 cells which include recombinant OvGM2AP . The minimum peptide length required was set to six residues and a minimum of two peptides were required to consider a protein as identified . The protein identification list was filtered at a False Discovery Rate below 1% . A synthetic peptide was generated following bioinformatics analysis of the OvGM2AP protein using OptimumAntigen design tool . The best peptide was selected on the basis of low host homology and high Antigenicity/Surface/Hydrophilicity index . The synthetic peptide corresponded to amino acid positions 62–72 of the full-length protein with the following sequence from N to C terminal ‘SSKSDGVKFTAEKS’ . This synthetic peptide was conjugated to Keyhole Limpet Hemocyanin ( KLH ) and used to immunize rabbits ( GenScript , USA ) ; after a secondary booster , immune serum was collected from rabbits and the antibodies were further purified from the immune serum by affinity chromatography using the synthetic peptide . Approximately 3 . 5 mg of antibodies was supplied with a titre greater than 1:64 , 000 , affinity purified from 2 rabbits . Total RNA was extracted from the different O . volvulus parasite stages using the Ambion RecoverAll Total Nucleic Acid Isolation kit ( Thermo Fisher Scientific , Belgium ) . The RNA was reverse-transcribed to cDNA using the iScript cDNA synthesis kit ( BioRad , USA ) . All primers were purchased from Integrated DNA Technologies ( IDT , Belgium ) . Primers were designed targeting OvGM2AP cDNA and amplifying a 165 bp product ( Forward primer 5’ GCCGAACAGCTCTGGAATTTG 3’ , Reverse primer 5’ TCGGTGACATGCGATCAGAC 3’ ) and these primers were used to amplify the OvGM2AP cDNA sequence from the parasite total cDNA as well as the genomic DNA . The designed primer set spans exon 5 to exon 6 ( S1 Fig ) . O . volvulus Glutaraldehyde-3-Phosphate Dehydrogenase ( OvGAPDH ) was used as a reference gene and a 192 bp fragment was also amplified from cDNA and genomic DNA preparations using the following primers; Forward primer 5’ GAAGGGTGGCGCTAAGAAAG 3’; Reverse primer 5’ GTTGTTGCATGTACGGTGGT 3’ . Amplification was carried out using the SensoQuest thermocycler ( Göttingen , Germany ) under the following conditions: 94 oC for 2 min , 1 cycle; 94 oC for 10 s , 55 oC for 10 s and 72 oC for 1 min , 40 cycles; 72 oC for 10 min . The reaction mix was prepared using the ReadyMix REDTaq PCR reaction mix ( Sigma ) according to manufacturer’s protocol . The PCR products were run on a 1% agarose gel stained with SYBR safe and visualized under UV light using the Gel Doc XR+ ( BioRad , USA ) . A western blot analysis of ESPs of different parasite stages was performed using anti-OvGM2AP peptide antibodies ( GenScript , USA ) . Approximately 10 μg of proteins were loaded onto SDS-PAGE gels and run at 200 V . The proteins were transferred to nitrocellulose membranes and blocked with 10% skimmed milk ( Régilait , France ) overnight at 4 oC followed by incubation with primary antibody at a dilution of 1:2000 for 1 h . After three changes of washing buffer ( PBS + 0 . 05% Tween 20 ) at 5 min interval each , goat anti-rabbit IgG ( whole molecule ) HRP conjugate ( Sigma ) was used as secondary antibody and incubated at a dilution of 1:5000 for 45 min at 37 oC . The membranes were again washed thrice with wash buffer , revealed by chemiluminescence using the ECL substrate and visualized using a C-Digit chemiluminescence scanner ( LI-COR , USA ) . IgG response to OvGM2AP was investigated by western blot and indirect ELISA using patient and control sera . For ELISA , optimal antigen concentration was determined by the checkerboard titration method . Maxisorp 96 well microtiter plates ( Nunc , Denmark ) were coated with 2 μg/ml of purified OvGM2AP overnight at 4 oC . Plates were washed thrice with wash buffer , 5 min between each wash and blocked with 10 mg/ml of Bovine Serum Albumin ( BSA ) for 1 h 30 min at 37 oC . The plates were washed as above and incubated with the various serum samples as primary antibody at a dilution of 1:2000 for 1 h at 37 oC . Following three rounds of washes at 5 min interval each , the plates were incubated with goat anti-human IgG ( Fc Specific ) peroxidase conjugate ( Sigma ) as secondary antibody at a dilution of 1:5000 for 1 h at 37 oC . After a final wash , 3 , 3’ , 5 , 5’ tetramethylbenzidine ( TMB ) was added as substrate for 10 min at 37 oC . The reactions were stopped with 2M sulfuric acid after which Optical Densities ( OD ) were read at 450 nm using the iMark microplate reader ( BIORAD , USA ) . All washes and antibody dilutions were done in wash buffer ( PBS+0 . 05%Tween-20 ) . IgG subclass responses were analyzed as described above with the exception of mouse anti-human IgG subclasses ( Sigma ) as secondary antibody and an additional incubation step of rabbit anti-mouse IgG peroxidase conjugate ( Sigma ) as tertiary antibody . For western blot , approximately 700 ng of purified OvGM2AP was transferred from SDS-PAGE gels to nitrocellulose membranes . The rest of the experimental procedures were carried as described for ELISA above except for the use of 10% skimmed milk ( Régilait , France ) for blocking and antibody dilution , the use of the Enhanced chemiluminescent ( ECL ) substrate and revelation using the chemiluminescence scanner . For the assessment of GM2AP activity of OvGM2AP , preparation of human glycosylated recombinant GM2AP with hexahistidine-tag was done as previously described [53] . The enzyme Hex A was purified from human placenta as described for the purification of sphingomyelinase [54] . [14C] GM2 was synthesized from its corresponding lyso-lipid , following published procedures [55] . This radiolabeled [14C] GM2 was incorporated into large unilamellar vesicles as described before [53 , 56] . Liposomes contained 20 mol% bis ( monoacylglycero ) phosphate ( BMP ) ( Sigma , Germany ) , 5 mol% cholesterol ( Sigma , Germany ) , 1 mol% [14C] GM2 and 1 , 2-Dioleoyl-sn-glycerol-3-phosphocholine ( DOPC ) ( Avanti polar lipids , Alabaster , USA ) as a host lipid in 20 mM sodium citrate buffer , pH 4 . 2 . Total lipid concentration was measured to be 50 mM . The actual liposomal activity assay of GM2AP with Hex A was carried out as previously described [53] . Briefly , 40 μl of liposome dispersion was mixed with 6 mU β hexosaminidase A and for BMP containing vesicles with 4 μg OvGM2AP or 4 μg recombinant human glycosylated GM2AP with a hexahistidine tag made up to 80 μl with 20 mM sodium citrate buffer pH 4 . 2 . The samples were incubated at 37°C for 30 min . Afterwards , the assay was put on ice and stopped by the addition of 20 μl chloroform/ methanol ( 1/1 , v/v ) . Quantification of the generated [14C] GM3 from [14C] GM2 was done by thin layer chromatography . The preparations were dried under a stream of nitrogen , re-dissolved with 20 μl chloroform/methanol ( 1/1 , v/v ) , vortexed and sonified for 15 min after which the solution of lipids was applied to a high-performance thin-layer chromatography plate ( Merck , Darmstadt , Germany ) . Lipids were separated in chloroform/methanol/0 . 22% CaCl2 ( 55/45/10 , v/v/v ) . Radioactive bands were visualized with a Bio Imaging Analyzer 1000 ( Fuji , Japan ) , and the quantification was performed with the image analysis software ‘‘Tina” ( Raytest , Staubenhardt , Germany ) . To assess if OvGM2AP has functional domains analogous to the human orthologue , we employed the use of the artificial fluorogenic substrate 4-methylumbelliferyl-2-acetamido-2-deoxy-β-D-glucopyranoside ( MUG ) ( Sigma , Germany ) and its sulfated derivative 4-methylumbelliferyl-2-acetamido-2-deoxy-6-sulfo-β-D-glucopyranoside ( MUGS ) ( Sigma , Germany ) in competitive inhibition with OvGM2AP . The assays were performed as previously described [57] . Briefly , a 40 μl reaction mix was constituted for each assay set comprising of 10 mM citrate buffer , pH 4 . 2 , 2 mM MUG or 2 x 10−5 M MUGS , 0 . 250 μg human β-hexosaminidase A ( R&D systems , USA ) , 6 μg BSA ( Sigma , Germany ) and varying amounts of OvGM2AP . The reaction was incubated for 30 min at 37 oC after which the fluorescence of the released 4-umbelliferone was determined following excitation and emission at 320 nm and 430 nm respectively using a Tecan scanner ( Infinite F200 Pro , Austria ) . Normality of distributions was assessed using a Shapiro-Wilk test . Normally distributed data are expressed as mean +/- standard deviation and were compared using parametric tests . For non-Gaussian distributions , data are expressed as median with interquartile ranges and were compared using non-parametric tests . Comparisons of more than two groups were made using a one-way analysis of variance ( ANOVA ) or a Kruskal-Wallis test ( with Dunn’s or Tukey’s correction for multiple comparisons ) for independent groups as appropriate . The discriminatory performance of total IgG , IgG1 , IgG2 , IgG3 and IgG4 was assessed using receiver operating curve ( ROC ) analyses . Areas under the ROC curves ( AUCs ) were evaluated using the trapezoid method . Standard errors of AUCs were calculated as previously described [58] . Exact confidence intervals for the AUCs were determined using a binomial approach . A p-value < 0 . 05 was considered statistically significant . Calculations were performed using SigmaPlot for Windows , version 12 . 5 ( Systat Software Inc . , Chicago , IL , USA ) . Diagnostic sensitivity and specificity as well as other diagnostic accuracy parameters were calculated as previously described [59] . Scatter plots were generated using Graph Pad Prism 5 . 0 ( La Jolla , CA , USA ) . 10 . 1371/journal . pntd . 0007591 . t001 No Organism Accession number Database 1 Onchocerca volvulus OVOC1952 WormBase Parasite 2 O . ochengi nOo . 2 . 0 . 1 . g11554 WormBase Parasite 3 Trichinella pseudospiralis T4C_8078 . 1 WormBase Parasite 4 Loa loa EN70_10408 WormBase Parasite 5 Caenorhabditis elegans C34E7 . 4 WormBase 6 Ascaris suum ASU_07674 WormBase Parasite 7 A . lumbricoides ALUE_0002125701 WormBase Parasite 8 Toxocara canis A0A183UTW5 WormBase Parasite 9 Dirofilaria immitis nDi . 2 . 2 . 2 . g07380 WormBase Parasite 10 Brugia malayi Bm2577 WormBase Parasite 11 Wuchereria bancrofti J9F8Z0 Uniprot 12 Homo sapiens P17900 Uniprot
The complete cDNA sequence coding for OvGM2AP was retrieved from NCBI non-redundant database by blasting the reverse complementary of the partial sequence referred to as OvL3 . C1 [60] . The full-length protein is assigned a WormBase ID of OVOC1952 . It contains 259 amino acids comprising 10 cysteine residues , 6 of which occupy the same positions in other species such as Homo sapiens , Loa loa , Trichinella pseudospiralis , and the free-living Caenorhabditis elegans ( Fig 1A ) . INTERPRO analysis of this protein revealed that it belongs to the superfamily of the GM2 activator protein and has a lipid binding domain . SUPERFAMILY , a member database of InterPro which is a library of profile hidden Markov models representing all proteins of known structure , was used to construct the entry . This same analysis holds true for orthologues of the protein in other nematode species and remains uncharacterized in all these species , except for T . pseudospiralis . Gene Ontology suggest it may be involved in biological processes such as nematode larval development , body morphogenesis , molting cycle , growth and locomotion . The degree of relatedness of OvGM2AP with orthologues in other nematodes depicts a conservation of the protein across species and further supports O . ochengi as the closest relative of O . volvulus ( Fig 1B ) . The protein sequence was found to contain a signal peptide ( amino acids 1–27 for OvGM2AP and 1–23 for human GM2AP ) in most orthologues . The amino acid sequences corresponding to the β-hexosaminidase A binding domain for the human GM2AP as well as the putative β-hexosaminidase A binding domain for OvGM2AP are illustrated in Fig 2A . Phyre2 , a structure prediction online tool not affiliated to the INTERPRO consortium , equally predicted OvGM2AP as a GM2AP with greater than 90% confidence . The complete list of hits obtained from the Phyre2 blast have been compiled in S2 Table . A model of the structure of the human GM2AP template is indicated in Fig 2B In terms of amino acid composition of the protein , serine was the most abundant , with 23 residues ( 8 . 9% ) . Other amino acids with structural implications present in the protein include Glycine ( 7 . 3% ) and Proline ( 5 . 8% ) . In order to determine the timing of OvGM2AP transcriptional expression during the various life cycle stages of the parasite , total RNA was purified from the larval stage 1 ( L1 ) or Mf , stage 2 ( L2 ) , stage 3 ( L3 ) , adult male ( AM ) as well as adult female ( AF ) stages and analyzed by RT-PCR using primers targeting OvGM2AP exon 5 ( forward primer ) and exon 6 ( reverse primer ) and generating a 165 bp product . Results obtained reveal that OvGM2AP mRNA is present in L1 , L2 , L3 , adult male ( AM ) and adult female ( AF ) stages of O . volvulus ( Fig 3A ) . To confirm that the detected bands of interest were not resulting from DNA contamination of the samples , the same primer pairs were also used to analyze O . volvulus genomic DNA ( gDNA ) . As shown in Fig 3A ( right panel ) , our primers detected signals with higher sizes , consistent with the genomic position of the targeted exons separated by intron 5 ( S1 Fig ) . The size difference between RNA and gDNA samples was also visible using O . volvulus GAPDH primers as control ( Fig 3A , lower panels ) , again consistent with the genomic positions of the targeted exons . The primer set designed to amplify OvGM2AP in O . volvulus samples was also successfully used to amplify the gene in O . ochengi species with exact precision as in O . volvulus further supporting the genetic closeness between the two species . In conclusion , the presence of the transcript in all the analyzed parasite stages suggests its permanent expression , which required confirmation at the translation level . In order to address this translational expression experimentally and since in silico analysis of OvGM2AP protein sequence suggests that it is a secreted protein , L3 , AM and AF O . volvulus worms were cultured in-vitro for 16 h and their corresponding ESPs were analyzed by western blotting using an anti-OvGM2AP peptide antibody . OvGM2AP was detected in 16 h in-vitro ESPs of third stage larvae ( L3 ) , AM and AF with the anti-OvGM2AP peptide antibodies but not with the pre-immune serum ( Fig 3B ) . This suggest that OvGM2AP is expressed and secreted by the parasite in its host at the tested life cycle stages and likely at all the different stages as revealed by RT-PCR analysis above . A constant and reliable source of recombinant antigen is required for in depth analysis of the OvGM2AP antigen . Attempts to produce soluble recombinant OvGM2AP in E . coli were only partially successful as mass spectrometry analysis of the purified protein showed the dominance of chaperones ( S2 Fig ) , we therefore switched to the insect cell baculovirus expression system . A recombinant virus encoding the OvGM2AP in fusion with a C-terminal 8His-tag under the control of the late pH promoter was generated by homologous recombination and was tested for its capacity to produce soluble OvGM2AP . This led to the production of the full-length protein including the TEV cleavage site and 8x His expected with a MW of 32 KDa ( Fig 3C ( iii ) ) . Western blot analysis of the clarified lysate from cells infected by the recombinant virus detected a protein of approximately 35 kDa that reacted with the serum of a rabbit immunized against an OvGM2AP peptide but not with the pre-immune serum ( Fig 3C ( i ) ) . As OvGM2AP is a secreted protein with a putative classical signal peptide , IMAC affinity purification of the histidine-tagged recombinant protein was assessed from cell lysates as well as from the culture medium . The latter purification strategy was successful and purification of OvGM2AP_8His from the culture medium using a single Ni-NTA affinity step yielded a pure protein ( Fig 3C ) with a concentration close to 1 mg per liter of culture . Mass spectrometry confirmed the identity of the purified protein with 23 unique peptides detected and a sequence coverage of 78 . 88% ( S3 Table ) . We could not detect any peptide corresponding to the N-terminal signal sequence of OvGM2AP as identified by sequence analysis . This suggests the peptide was cleaved and could constitute the signal peptide sequence of the protein . We next investigated the antigenic potential of the insect cell expressed and purified recombinant OvGM2AP . We first asked if the protein , new for analysis in Onchocerca , could distinguish between onchocerciasis patients and onchocerciasis non-infected individuals . When a 14 sera pool of O . volvulus Serum ( OVS ) and Hypoendemic Serum ( HES ) control , as well as a 3 sera pool of European serum control ( ESC ) were employed for western blot analysis of 700 ng of purified OvGM2AP , OvGM2AP was found to react specifically with patient sera ( OVS ) and not with HES and ESC controls ( Fig 3D ) . The presence of other bands of higher molecular weight suggests a possible polymerization of the native protein , as also observed with the in vitro ESP analysis above ( Fig 3B ) . The Homo sapiens GM2AP is known to be glycosylated at asparagine 63 [61] . There is experimental evidence of glycosylation of the C . elegans orthologous protein at asparagine 85 [62] . It therefore appears that the glycosylated asparagine amino acid residue differs across species and also between different members of the family . We herein analyzed the recombinant OvGM2AP produced in insect cell for the possible presence of glycosylation . To achieve this , Peptide:N-glycosidase F ( PNGase F ) was used to digest the recombinant OvGM2AP and the digestion product was resolved by SDS-PAGE . As shown ( see arrow head ) in Fig 4A ( i ) , a band shift was observed between the digested and undigested purified protein . This band shift was confirmed by western blot analysis using the anti-OvGM2AP peptide antibody ( Fig 4A ( ii ) ) . The band corresponding to the enzyme used in the digestion ( see arrow , Fig 4A ( i ) ) failed to show a signal by western blot , further testifying to the specificity of our anti-OvGM2AP peptide antibody . To verify if the shift was associated with enzymatic deglycosylation , mass spectrometry analysis was performed on peptides obtained from trypsin digestion of PNGase F treated and untreated OvGM2AP samples . The asparagine at position 173 in the native recombinant OvGM2AP was found to be deamidated to aspartate in the digested protein ( Fig 4B ) , a consequence of PNGase F digestion of N-linked glycans from glycoproteins . Two different glycosylation sites were observed by mass spectrometry with 100% and 86% confidence corresponding to the peptide “NISLRICLPTK” and peptide “NLEPGKYKNISLR” respectively . This PNGase F treatment therefore suggests that the purified recombinant OvGM2AP is glycosylated predominantly at Asn 173 but with the possibility of a second glycosylation site at Asn 165 . The ion exchange chromatogram corresponding to the deamidation peaks of the identified glycosylated peptides is provided in S3 Fig . Altogether , this supports the glycosylation of the protein at asparagine 173 following the pattern N-X-S where X is any amino acid except proline . As an attempt to understand the immune response to OvGM2AP , total IgG responses were measured by ELISA in infected and non-infected individuals . Results obtained indicate , consistent with the western blot analysis reported above ( Fig 3D ) , a discriminatory immune response to OvGM2AP between infected and non-infected individuals . The mean OD450nm for patient sera ( OVS ) was significantly different ( P<0 . 05 ) from that of normal African sera ( HES ) and European sera ( ESC ) with p-values lower than 0 . 001 for both HES and ESC ( Fig 5A ) . The area under the ROC curve ( AUC ) was found to be high , with a value of 0 . 9863 , and a p-value < 0 . 0001 ( Table 1; S4 Fig ) indicating both high sensitivity and specificity . We further investigated IgG subclass responses to the recombinant OvGM2AP by ELISA . All IgG subclasses ( with the exception of IgG3 ) were detected at significantly different levels between infected individuals ( OVS ) and uninfected African Controls ( HES ) ( Fig 5B–5E ) . The IgG4 subclass was found to have a higher AUC value ( 0 . 9451 ) compared to the other subclasses with a p-value < 0 . 0001 ( Table 1 ) . Diagnostic accuracy parameters investigated for the IgG and IgG subclass responses also revealed IgG4 as the favorite ( Table 2 ) . In a bid to assess OvGM2AP as a possible marker of a decrease of infection related to ivermectin treatment , we decided to investigate the correlation between the humoral immune response and rounds of ivermectin ( IVM ) intake . To this end , sera from Bandjoun in the Western region of Cameroon where onchocerciasis is almost eliminated [46] were grouped ( 10 samples per group ) according to the number of rounds of ivermectin intake ( expressed as years of treatment ) and analyzed by ELISA using OvGM2AP as a bait . As shown in Fig 6A , results obtained suggest a negative correlation between the anti-OvGM2AP response and rounds of IVM intake . Consistently , samples from this site were compared to those from Kombone in the South-Western region where onchocerciasis remains actively endemic . In the analysis , sera from Rwanda where this disease is hypo-endemic ( HES ) were used as negative controls together with European sera ( ESC ) . As shown in Fig 6B , OvGM2AP significantly discriminated between sera from the endemic region of Kombone ( OVS ) and the hypoendemic region of Bandjoun . This suggests that monitoring anti-OvGM2AP immune response would allow the discrimination of endemic and hypoendemic regions as well as the evaluation of the state of infection following ivermectin treatment . In order to investigate a possible cross-reaction with OvGM2AP in individuals infected with other nematodes , we analyzed sera from patients infected with Loa loa , Brugia malayi , Wuchereria bancrofti , Mansonella perstans and Ascaris lumbricoides . With this in mind , OvGM2AP was used to coat ELISA plates as earlier described and IgG response was measured in L . loa infected serum ( LLS ) , W . bancrofti infected serum ( WBS ) , B . malayi infected serum ( BMS ) , M . perstans infected serum ( MPS ) , A . lumbricoides infected serum ( ALS ) and compared with OVS and controls . Results obtained indicated a very strong response from related nematodes ( Fig 7 ) . This suggests consistently with the high sequence identity shown above ( Fig 1A ) , that epitopes are common to Ov and Ll GM2AP , as well as the other nematodes . This is a drawback for the use of OvGM2AP in the specific diagnosis of the disease . The ability of OvGM2AP to mediate the hydrolysis of GM2 by β-hexosaminidase A was investigated in an in vitro liposomal assay using negatively charged , BMP-containing liposomes . Liposomal composition was Cholesterol ( 5 mol% ) , BMP ( 20 mol% ) and [14C] GM2 ( 1 mol% ) , made up to 100 mol% by DOPC . Turnover of [14C] GM2 in the liposomal activity assay was stimulated by human recombinant GM2AP or tauridesoxicolate ( TDC ) ( 50 μg ) , both used as positive control . No hydrolysis of [14C] GM2 was detected in the presence of OvGM2AP ( Fig 8A ) . The human GM2AP has been reported to competitively inhibit the degradation of the artificial fluorogenic MUGS substrate but not the neutral MUG substrate . In a bid to understand if a parasitic GM2AP like OvGM2AP could share the same characteristics like the human GM2AP , we employed MUG and MUGS in a competitive inhibition assay with OvGM2AP . As shown in Fig 8B , OvGM2AP was found to competitively inhibit the degradation of MUG , but not MUGS as previously reported for the human GM2AP [63] .
Parasite genomics has evolved drastically within the last decade with more genomes being sequenced . The recently published genome of Onchocerca revealed a 44% of O . volvulus genes coding for proteins with no predicted function [21] . In order to deeply understand the biology of the parasite , these unknowns should be solved . In this study , we attempted to characterize one such uncharacterized protein belonging to the O . volvulus ESP family which was identified by INTERPRO analysis to be a member of GM2 activator proteins ( GM2APs ) . The canonical GM2APs are accessory glycoproteins required for the in vivo degradation of ganglioside GM2 to GM3 in the lysosomal compartments by β-hexosaminidase A [64] . The latter can cleave glycolipid substrates on membrane surfaces only if they extend far enough into the aqueous phase . In the absence of detergents , the degradation of ganglioside GM2 occurs only in the presence of the GM2 activator protein [41] . GM2AP has also been shown to act as a lipid transfer protein [37 , 38] . Although INTERPRO analysis identified Ov28CRP as a GM2AP , its size is a little bigger and the amino acid identity is very low , suggesting the O . volvulus protein could be a GM2AP member with novel functions . Moreover , we have provided experimental evidence of its secretion in 16 h in vitro ESPs collected from L3 , adult male and adult female stages , supporting the presence of the protein in a parasitic ESP in vivo . Trichinella pseudospiralis GM2AP has been reported to be secreted following its expression in yeast cells . This parasite GM2AP did not , however , facilitate degradation of GM2 ganglioside by N-acetylhexosaminidase A , although it did inhibit phospholipase D ( PLD ) activity . Lack of the former activity might be explained by the absence of a domain implicated in binding to hexosaminidase [65] . In this study , we also did not observe hydrolysis of GM2 to GM3 by OvGM2AP , supporting a trend that this activity might be lost in parasitic species . Additionally , we observed competitive inhibition of MUG degradation with OvGM2AP ( Fig 8B ) and not MUGS as previously reported for the human GM2AP [63] . The competitive inhibition of MUGS with the human GM2AP was proposed to be as a result of the binding of GM2AP/MUGS to a site on the α subunit of β-hexosaminidase A ( Hex A ) . However , a recent model of GM2AP/GM2/Hex A was provided , predicting the binding of GM2AP to a site on the β subunit of HexA while it presents the GM2 to a positively charged site on the α-subunit which interacts with the negatively charged sialic acid of GM2 [66] . The inhibition of degradation of MUG which is a neutral substrate by OvGM2AP indicates the possible lack of a positively charged site in OvGM2AP that binds to the negatively charged MUGS substrate . This is a first indication of the protein playing an opposite role to that previously described for the human protein and could therefore be conferring some advantages in supporting a parasitic lifestyle . Another possible function of the GM2AP with potential significance in host-parasite interactions is phospholipase D ( PLD ) inhibition . PLD is involved in the signaling pathway , hydrolyzing the phosphodiester bond of the glycerolipid phosphatidylcholine to second messenger phosphatidic acid ( PA ) [67] . PLD and its second messenger PA have been implicated in a variety of biological processes including signal transduction and anti-apoptotic signaling , phagocytosis , exocytosis/secretion , as well as chemotaxis for neutrophils and cancer cells [68–73] . The E . coli expressed non-glycosylated human GM2AP can be recaptured from the extracellular medium into the cell [74] and this was reported to be mediated mainly through the mannose-6-phosphate pathway [75] . Investigating the ability of OvGM2AP to be internalized by immune cells will throw light into its possible immunomodulatory functions . It was also demonstrated that the binding abilities of the GM2AP are altered by the presence of a His-tag [76] . These observations provide critical information to take into account in our future biochemical characterization of the O . volvulus His-tagged recombinant OvGM2AP . In addition , and contrary to data on the human GM2AP , the nematode homologue failed to inhibit platelet activating factor-induced calcium mobilization in neutrophils , but actually enhanced mediator-induced chemotaxis [65] . As OvGM2AP is secreted , its presence in key parasite stages can be useful in the development of an antigen capture test as this would make it possible to detect the presence of L3s , adults and microfilariae in infected individuals thereby increasing its scope in detecting both pre-patent and active infection . While OvGM2AP functional elucidation remains a topic for a separate investigation , we have been able to characterize anti-OvGM2AP humoral immune responses using the recombinant protein expressed in SF21 insect cells . Our choice of insect cells in which posttranslational modifications are possible was dictated by a gain of knowledge of the immune response to a fully functional molecule . Analysis of the total IgG response to OvGM2AP revealed it as discriminant between infected and non-infected individuals . We furthermore analyzed the immune response to OvGM2AP in patients who have been administered different rounds of Ivermectin and found a statistically significant decline in IgG response with increased rounds of drug administration ( Fig 6A ) . This finding is useful in the context of the assessment of onchocerciasis control programs for which skin snip Ov16 based test are still being widely used for evaluation of the CDTI program in the definition of treatment endpoints [77] . Limitations of both tests have called for the search of new diagnostic antigens and thus allowing the design of antigen cocktails that may solve the present difficulties in obtaining high specificity without a corresponding drop in sensitivity [18] . The use of OvGM2AP as a constituent of antigen cocktails to determine ivermectin treatment endpoints might improve on its major drawback which is its cross-reactivity with related nematode infections , limiting its analytical specificity . Helminth parasites are known to modulate host immune responses to establish their long term stay in an immunocompetent host . Individuals with high parasite load are often asymptomatic and present only mild pathology . This is largely due to the fact that the immune response to helminth infection is less severe and regulatory [78] . Indeed , the immune response is usually TH2 biased , with a TH2 response shift either towards immunosuppression , immuno-tolerance or modified TH2 response [79] . The term modified TH2 response stands for a downplay of the downstream effects of normal TH2 responses resulting to an increase in non-complement fixing IgG4 and IL-10 [80 , 81] . In our results , we did not see a significant difference between the IgG4 subclass and the other IgG subclasses amongst patients ( as reflected by the OD values ( Fig 5 ) ) . However , the IgG4 subclass could better differentiate between Mf positive individuals ( OVS ) and uninfected African controls ( HES ) as indicated from the ROC curve . This could therefore indicate a bias towards IgG4 in response to OvGM2AP . In conclusion , during the course of this study , we have established that OvGM2AP ( i ) is an ESP of O . volvulus , ( ii ) is immunogenic and ( iii ) has characteristics compatible with its use for disease assessment following ivermectin treatment of human onchocerciasis . In this regard , the most important characteristic of OvGM2AP is its ability to discriminate between individuals being treated and those already treated: it therefore holds some potential in certification of disease elimination . Finally , the ongoing elucidation of OvGM2AP function and structure should provide valuable insights into host-parasite interactions as well as better understanding of the nematode biology .
|
Human onchocerciasis is a neglected tropical disease affecting millions in endemic tropical countries and is the world’s second leading cause of infectious blindness . Onchocerca volvulus , the causative agent of the disease is a tissue dwelling nematode diagnosed by identification of microfilaria larval stages in skin snip biopsies . The development of new tools for the management of the disease requires the characterization of more parasite antigens as about 44% of genes have no predicted function . In this research , efforts were made to characterize one such novel secretory protein of the parasite . We report that it is a putative orthologue of the human GM2 activator protein and evaluated its immune response in the context of disease diagnosis . Although we observed cross-reaction with Loa loa and other related nematodes , our data still supports further investigations towards a possible function of the protein in immunomodulation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"onchocerca",
"volvulus",
"helminths",
"purification",
"techniques",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"animals",
"onchocerca",
"nematode",
"infections",
"neglected",
"tropical",
"diseases",
"onchocerciasis",
"immunologic",
"techniques",
"protein",
"purification",
"research",
"and",
"analysis",
"methods",
"immunoassays",
"recombinant",
"protein",
"purification",
"immune",
"response",
"helminth",
"infections",
"eukaryota",
"nematoda",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2019
|
Identification and characterization of the Onchocerca volvulus Excretory Secretory Product Ov28CRP, a putative GM2 activator protein
|
Individuals with high intensity of Loa loa are at risk of developing serious adverse events ( SAEs ) post treatment with ivermectin . These SAEs have remained unclear and a programmatic impediment to the advancement of community directed treatment with ivermectin . The pathogenesis of these SAEs following ivermectin has never been investigated experimentally . The Loa/baboon ( Papio anubis ) model can be used to investigate the pathogenesis of Loa-associated encephalopathy following ivermectin treatment in humans . 12 baboons with microfilarial loads > 8 , 000mf/mL of blood were randomised into four groups: Group 1 ( control group receiving no drug ) , Group 2 receiving ivermectin ( IVM ) alone , Group 3 receiving ivermectin plus aspirin ( IVM + ASA ) , and Group 4 receiving ivermectin plus prednisone ( IVM + PSE ) . Blood samples collected before treatment and at Day 5 , 7 or 10 post treatment , were analysed for parasitological , hematological and biochemical parameters using standard techniques . Clinical monitoring of animals for side effects took place every 6 hours post treatment until autopsy . At autopsy free fluids and a large number of standard organs were collected , examined and tissues fixed in 10% buffered formalin and processed for standard haematoxylin-eosin staining and specific immunocytochemical staining . Mf counts dropped significantly ( p<0 . 05 ) in all animals following ivermectin treatment with reductions as high as ( 89 . 9% ) recorded; while no significant drop was observed in the control animals . Apart from haemoglobin ( Hb ) levels which recorded a significant ( p = 0 . 028 ) drop post treatment , all other haematological and biochemical parameters did not show any significant changes ( p>0 . 05 ) . All animals became withdrawn 48 hours after IVM administration . All treated animals recorded clinical manifestations including rashes , itching , diarrhoea , conjunctival haemorrhages , lymph node enlargement , pinkish ears , swollen face and restlessness; one animal died 5 hours after IVM administration . Macroscopic changes in post-mortem tissues observed comprised haemorrhages in the brain , lungs , heart , which seen in all groups given ivermectin but not in the untreated animals . Microscopically , the major cellular changes seen , which were present in all the ivermectin treated animals included microfilariae in varying degrees of degeneration in small vessels . These were frequently associated with fibrin deposition , endothelial changes including damage to the integrity of the blood vessel and the presence of extravascular erythrocytes ( haemorrhages ) . There was an increased presence of eosinophils and other chronic inflammatory types in certain tissues and organs , often in large numbers and associated with microfilarial destruction . Highly vascularized organs like the brain , heart , lungs and kidneys were observed to have more microfilariae in tissue sections . The number of mf seen in the brain and kidneys of animals administered IVM alone tripled that of control animals . Co-administration of IVM + PSE caused a greater increase in mf in the brain and kidneys while the reverse was noticed with the co-administration of IVM + ASA . The treatment of Loa hyper-microfilaraemic individuals with ivermectin produces a clinical spectrum that parallels that seen in Loa hyper-microfilaraemic humans treated with ivermectin . The utilization of this experimental model can contribute to the improved management of the adverse responses in humans .
Loa loa is a parasitic filarial nematode of humans , restricted to the rainforest and forest fringes of West and Central Africa [1] , causing the relatively well-tolerated disease-loiasis . Loiasis has two very characteristic clinical features: Calabar swellings , corresponding to an angioedema of allergic nature [2] and the passage of an adult worm under the conjunctiva ( eyeworm ) . Other complications , though rare , include nephropathy [3] , cardiomyopathy [4] , retinopathy [5] , arthritis [6] , lymphangitis [7] , peripheral neuropathy [8] and encephalopathy [9] . Loiasis is an infection of public health importance , essentially not because of its own clinical manifestations , but because it is associated with serious adverse events ( SAEs ) following the administration of ivermectin . These SAEs have a negative impact on the ongoing control of onchocerciasis and lymphatic filariasis with ivermectin distribution in areas of co-endemicity with Loa . The occurrence of Loa-associated SAEs with programs for these two major filariae ( onchocerciasis and lymphatic filariasis ) in L . loa endemic areas has been increasingly reported over the past decade . Clinical problems seen include severely disabling and potentially fatal , encephalopathy as well as other permanent clinical changes . Current observations have indicated that the risk of developing marked or serious reactions is significantly higher when the L . loa loads exceeds 8 , 000mf/mL; the severity of these reactions becomes more common in patients with the risk of these happening and being severe much higher in patients with > 30 , 000 mf/mL , with the risk of problems being very high with loads above with > 50 , 000mf/mL [10–13] . Although the epidemiological mapping and clinical description of these Loa-associated SAEs is now well known , the pathogenesis and the optimal approaches to treatment still remain obscure . An increase in the presence of L . loa in the brain tissue has been proposed as an important feature of the condition , and a central role for vascular pathology in the adverse reaction also proposed [14 , 15] . Understanding the mechanism of pathogenesis of the post-ivermectin events in these heavily infected individuals is an important step towards the development of better treatments and responses to this unfortunate pathology . The lack of data from human cases and especially the lack of autopsy material have hindered the progress in the treatment and prevention of this condition . The previous lack of a relevant experimental model also contributed to the lack of solid information on these adverse reactions and delay in developing suitable clinical treatment protocols . A better understanding of the pathogenesis of this post-treatment condition is urgently needed . After several consultation meetings held by the Scientific working Group on SAE in L . loa endemic countries , [16] , it was strongly recommended that a non-human primate ( NHP ) animal model of Loa encephalopathy be developed . Non-human primates ( NHP ) are therefore considered the best models for the much needed investigation in human loiasis , especially as suitable in vitro models for loiasis have not yet been developed , and indeed such artificial models are not easily extrapolated to humans [17–19] . Although the drill is an excellent experimental host , there are ethical concerns with using this protected animal for research , and it is no longer used in biomedicine . The use of Patas monkeys also is limited as the parasite behaves differently from the same way that it does in the more human-like drill [20] . As recently described , the baboon ( P . anubis ) is a potentially useful model for studying the mechanisms behind the SAEs that develop in L . loa infected people as the parasite in this animal behaves essentially in the same way as it does in the drill . The use of baboons in biomedical research is accepted by the International Union for Conservation of Nature ( IUCN ) [21] and so Wanji et al developed a hyper-microfilaraemic Loa/baboon ( Papio anubis ) experimental model and characterized it parasitologically , haematologically and biochemically [22] . In this model microfilariaemia increases steadily in all infected animals and reaches a peak at 18 months post infection ( MPI ) ; by 10 MPI >70% of animals have mf > 8 , 000 mf/mL , at 18 MPI >70% of animals have mf >30 , 000mf/mL and 50% of animals have mf >50 , 000mf/mL . In this study , the three most significant alterations seen were: increased eosinophil , creatinine and gamma-GT levels . An intriguing question that now emerges is whether these specific alterations bear any role in the development of the post ivermectin SAEs . It also remains to be demonstrated whether if these Loa hyper-microfilaraemic baboons are treated with ivermectin-IVM ( i ) There will be a drastic drop in microfilariae load ? ( ii ) Treatment will lead to serious clinical manifestations ? ( iii ) There will be changes in blood chemistry and haematological parameters ? ( iv ) Microfilariae will be seen in the different tissues and their distribution ? ( v ) Which lesions will be seen and what are their characteristics ( vi ) and if the co-administration of ivermectin with either aspirin or prednisone will affect points i-v above . Therefore , this present study was carried out to treat Loa hyper-microfilaraemic animals such as to validate the model as a surrogate for the human condition , and to provide answers to the questions above .
The acquisition , care and ethical concerns on the use of these animals have already been described in the preceding paper [22] . The animals for this study were gotten from the 15 baboons that had been characterised parasitologically , biochemically and haematologically in the preceding paper [22] . Baboons to be included in the trials were chosen amongst the experimentally infected 15 animals . The inclusion criterion for animals to enter the drug treatment phase was a microscopically confirmed microfilariaemia of >8 , 000mf/mL whilst the exclusion criterion was a microfilariaemia of <8 , 000mf/mL . Ultimately , 12 infected splenectomised baboons were included in this phase of the study . The 12 baboons were randomised into 4 arms , with each treated group consisting of three animals . Eligible baboons were randomly assigned to the different treatment arms using a computer generated randomisation list . The randomisation was performed by a statistician not otherwise engaged in the study . The initial treatment allocation and the number of days animals were to be monitored were concealed but the subsequent administration of medication ( s ) was open label . Animals were placed into three different experiments based on the number of days animals were to be monitored; animals in experiment 1 were monitored for 5 days , those in experiment 2 for 7 days and those in experiment 3 for 10 days . After the randomization into treatment groups , the study was designed according to the flow diagram shown in Fig 1 . Ten mL venous blood samples were collected under anesthesia ( 0 . 1 mL ketamine ) and aseptic conditions from the femoral vein of each animal for parasitological ( L . loa ) , haematological and biochemical analyses on Day 0 before treatment and on Days 5 , 7 or 10 post treatment depending on the animals experimental group by a licensed veterinarian . Calibrated thick blood smears were prepared by spreading a 50 μL venous blood sample from a 75 μL non-heparinised capillary tube , onto a clean slide over an area of 1 . 5 cm x 2 . 5 cm for both the pre-treatment and post treatment samples . The haematological parameters assessed were haemoglobin ( Hb ) , red blood cell count ( RBC ) , total white blood cell count ( WBC ) and white cell differential count in which neutrophils , eosinophils , basophils , monocytes and lymphocyte counts were measured . These analyses were performed on the pre- and post-treatment samples . The biochemical parameters quantified were grouped into the liver enzymes consisting of serum glutamate-pyruvate transaminase ( SGPT ) , serum glutamate-oxaloacetate transaminase ( SGOT ) and serum γ-glutamyl transferase ( γ-GT ) ; blood biochemistry compounds comprising of creatinine and glucose; and blood elemental biochemistry comprising of calcium and potassium . The biochemical parameters were quantified using spectrophotometric kits obtained from HUMAN ( www . human . de , Germany ) as per the manufacturer’s instructions . The parasitological , haematological and biochemical analyses were performed on the pre- and post-treatment samples as described previously in Wanji et al . , [22] . The ivermectin used for this study was provided by the Mectizan Donation Program ( MDP ) , and the aspirin ( ASA ) and prednisone ( PSE ) used in the study acquired commercially from Bayer and Ranbaxy laboratories Ltd . , respectively . The animals were administered drugs via the oral route by a licensed veterinarian . Group 1 animal served as controls and received no drugs . Group 2 animals ( IVM ) received ivermectin alone . Group 3 animals ( IVM+ASA ) received IVM and 3 days after IVM administration , they received aspirin ( 1 tablet of 500mg 2 times a day ) for two days; Group 4 animals ( IVM+PSE ) received IVM and 3 days after IVM administration , they received PSE ( 4 tablets of 5mg 2 times a day ) for two days . The standard doses of ivermectin ( 150μg/kg of body weight ) , 500mg of ASA and 5mg of PSE were used for this study . Clinical surveillance was carried out for up to 10 days . Following treatment , animals were closely monitored every day on a six hourly basis . Animals were assessed for general agility , meningeal signs , discharges , papillary reflex , diarrhea , weakness , conjunctival haemorrhage , dyspnea , itching ( excoriations ) , muscle aches , restlessness , gland pain and gland tenderness , presence of rashes , pinkish ears , joint pain and swollen face . The agility and playing habits of the monkeys were assessed by the animal caretaker who was familiar with the normal behaviour of these baboons . Rashes were considered absent when none were found on the body of the animal , mild when < 3% , moderate when 3–10% , severe when 11–50% and unbearable when > 50% , of the body surface was affected . Itching was considered absent when no excoriations were found on the body of the animal , to be mild when < 3% of the body surface had excoriations , moderate when 3–10% , severe when 11–50% , and unbearable when > 50% of the body surface had excoriations . Gland pain/tenderness was assessed by palpating collections of lymph nodes in the neck , axillae and inguinal regions , and were considered as absent if no enlarged nodes were found; mild when the enlarged nodes were localized i . e . 1–2 enlarged nodes were found in one body area; moderate if 3–5 enlarged nodes were found in one body area; severe when 3–7 enlarged lymph nodes were generalized i . e . found in 2 or more body areas and unbearable when > 10 enlarged lymph nodes were generalized . Restlessness was assessed by counting the number of times the animals could not sit still; restlessness was considered absent when the animals sat still all the time , mild when it could not be still for less than 5 times , moderate when it could not be still for about 8–15 times , severe when it could not be still for 15 times and unbearable when the animal could not remain sitting still . Vital signs like rectal temperature , pulse rate ( PR ) and respiratory rate ( RR ) were also recorded . Animals in Experiment 1 were monitored every 6 hours for 5 days for a period of 120 mins , those in Experiment 2 every 6 hours for 7 days for a period of 168 mins while those in Experiment 3 were monitored every 6 hours for 10 days for 240 mins before autopsy . The animals were euthanized by the administration of 10 mL of 5mg ketamine ( Imalgene , Merial , France ) and autopsies performed immediately . Gross organ and pathophysiologic abnormalities were observed during necropsy . Organs observed were the skin , brain , liver , lungs , diaphragm , pancreas , kidneys , heart , digestive tract and lymph nodes . These organs were collected and fixed in 10% formol ( 3% formaldehyde ) . Following paraffin embedding and sectioning ( 5 μm ) the tissues were stained with a range of histochemical ( haematoxylin-eosin ) and immunochemical reagents ( REF-350 was used to identify thrombi and UCH-ICC was used to identify the presence of macrophage cell lines ) using standard immunocytochemical procedures from the Department of Pathology/Histology of the Michigan State University , Michigan , U . S . A . Slides with stained tissues of the brain , liver , lungs , skin , diaphragm , pancreas , kidneys , heart , digestive tract and lymph nodes were used to estimate Loa mf numbers and also analysed for pathologic changes . Ruled out papers of 50mm2 and 12 . 5mm2 were used to carve out areas on the prepared slides for organs with large surface areas ( e . g . brain , kidney , lungs , etc . ) and those with small surface areas ( e . g . skin , lymph nodes , pancreas , etc . ) , respectively . The slides were rapidly focused at x10 to trace out the borders of the marked area and to check for Loa mf , after which the x40 objective was used to confirm the presence of Loa parasites and estimate the Loa mf numbers . After the counts had been made , the distribution of mf/25mm2 in all organs was calculated . The histopathological tissues were observed and interpreted by two trained histopathologists ( CDM , DA ) . Normal ranges for various vital signs ( temperature , respiratory rate ( RR ) and pulse rate ( PR ) in baboons were gotten from the Association of Primate Veterinarians Primate Formulary ( 1999 ) for baboons housed individually in large custom cages . The data were entered into Epi-Info version 3 . 5 . 3 ( C . D . C . Atlanta , GA , U . S . A . ) and analysed using the Software Package SPSS version 20 . Graphs were drawn using the Graph Pad Prism version 5 for windows , Graphpad Software , San Diego California , U . S . A . Descriptive statistical analyses were performed to compute the mean , median and standard deviations of Loa microfilarial counts , haematological and biochemical parameters in the animals pre- and post-treatment . The percentage reduction of peripheral mf following treatment in the different baboons was calculated using the formula below: %reductionofmf=mfbeforetreatment-mfbeforetreatmentx100mfaftertreatment The Pearson’s Chi square test was used to test for significant differences in percentage reductions of mf between groups . The Wilcoxon signed rank test was used to test for any significant differences in mf counts , haematological and biochemical parameters before and after treatment . All clinical manifestations observed were described based on scores defined as: 0 = normal/absent/no alteration , 5 = mild , 10 = moderate , 15 = severe and 20 = unbearable . The Kruskal and Wallis test was used to test for significant differences in vital signs . Bar charts were used to represent the mean scores of clinical signs in the different treatment groups . The distribution of mf/25mm2 in the different organs for animals in the different treatment arms were represented using bar charts; the Kruskal and Wallis test was used to test for significant differences in the distribution of tissue mfs . A description of the different lesions observed at the gross and histological levels were done . The mf found in the different tissues of Bab 08 was considered under the category of IVM only , because it took only this drug and died 5 hours after . All tests were performed at 5% significance level .
The pre-treatment Loa microfilaraemia load ranged from 19 , 800–124 , 700 mf/mL; median 39 , 500 mf/mL of blood and the post-treatment microfilarial load ranged from 160–34 , 660mf/mL; median 12 , 900 mf/mL of blood . The untreated animals showed an overall percentage fluctuation of 3 . 4% over the 10 days period monitored . After 5 days of monitoring , these animals showed a reduction of 22 . 6% whereas the animals monitored for 7 and 10 days showed increases in the mf numbers of 8 . 8% and 3 . 5% , respectively from day 0 ( Fig 2A ) . For all animals administered IVM ( IVM alone , IVM+ASA and IVM+PSE ) , the overall reduction in mf load after ivermectin treatment was 75 . 7% . Animals monitored for 7 days recorded the highest ( 96 . 9% ) reduction in mf loads following treatment whereas those monitored for 5 days recorded the lowest ( 59 . 5% ) of the reductions seen in mf loads with treated animals ( Fig 2 ) . There was a significant difference ( p = 0 . 02 ) in percentage reduction in animals treated at different time points . For each treated animal , there was a significant drop in microfilaraemia p<0 . 05 , Table 1 . Bab 05 ( with an initial Loa mf load of 10 , 080 mf/mL ) recorded the highest ( 98 . 4% ) reduction in mf numbers while Bab 02 ( with an initial Loa mf load of 79 , 660 mf/mL ) recorded the lowest ( 61 . 8% ) reduction in mf numbers after treatment with IVM . The percentage reduction in individual animals treated with IVM is shown in Table 1 . With respect to treatment groups , the highest percentage reduction ( 89 . 9% ) in peripheral mf loads was recorded in animals that were administered IVM alone ( Fig 3 ) . There was a significant difference ( p = 0 . 03 ) in percentage reduction in animals in the different treatment arms administered ivermectin . Amongst all the haematological parameters monitored there was a significant difference between the pre- and post-haematological parameters when looked globally ( p = 0 . 028 ) . Haemoglobin values before treatment ranged from 12–16 g/ dL; median 14 . 7 g/dL while the post treatment values ranged from 12–16 g/dL; median 12 . 8 g/dL . However , the Hb values pre- and post-treatment remained the same ( 12g/dL ) in animals which received IVM + PSE while for animals that received IVM alone and IVM + ASA reduced post treatment by 3 g/dL and 1 g/dL , respectively . However , in untreated animals increases were noticed in the RBC numbers and decreases noticed in the WBC , neutrophil , mononuclear numbers before and after treatment , significant differences were not noticed in the values for these parameters pre and post treatment between the different treatment groups , ( p>0 . 05 ) ( S1 Table ) . Although significant differences were not recorded ( p>0 . 05 ) between the different treatment groups , IVM administered alone caused increases in the values pre and post treatment for WBC , eosinophils and mononuclear lymphocytes while a decrease in values pre and post treatment was noticed in neutrophil numbers ( S1 Table ) . IVM +ASA co-administration caused increases in the values pre and post treatment for neutrophils and decreases in RBC , WBC , eosinophils , mononuclear lymphocytes , although no significant differences ( p>0 . 05 ) were recorded in these values between the different treatment groups ( S1 Table ) . IVM + PSE co-administration caused increases in the values pre and post treatment for RBC and decreases in WBC , neutrophils eosinophils and mononuclear lymphocytes , although no significant differences ( p>0 . 05 ) were recorded in these values between the different treatment groups ( S1 Table ) . The different biochemical parameters monitored did not show any significant changes in values pre and post treatment ( p>0 . 05 ) . The changes in all biochemical parameters before and after treatment are shown in S2 Table . In untreated animals though increases were noticed in the SGPT , SGOT , γ-GT , creatinine and glucose levels numbers while decreases were noticed in the calcium and potassium numbers before and after treatment , significant differences were not noticed in the values for these parameters pre and post treatment , between the different treatment groups ( p>0 . 05 ) ( S2 Table ) . Although significant differences were not recorded ( p>0 . 05 ) between the different treatment groups , IVM administered alone caused increases in the values pre and post treatment for SGOT , glucose and calcium , while decreases in values pre and post treatment were noticed in SGPT , γ-GT , creatinine and potassium ( S2 Table ) . IVM +ASA co-administration caused increases in the values pre and post treatment for γ-GT , no changes in calcium levels and decreases in SGPT , SGOT , creatinine , glucose and potassium although no significant differences ( p>0 . 05 ) were recorded in these values between the different treatment groups ( S2 Table ) . IVM + PSE co-administration caused increases in SGPT , SGOT , γ-GT and glucose levels and decreases in creatinine , calcium and potassium levels although no significant differences ( p>0 . 05 ) were recorded in these values between the different treatment groups ( S2 Table ) . In all animals administered IVM at a standard dose of 150μg/kg on Day 0 , they did not show any clinical manifestations , however , Bab 08 ( with an initial Loa mf load of 124 , 700mf/mL ) became unbearably restless , lost appetite and died 5 hours after treatment . On the second day of observation post treatment , the following clinical manifestations being observed: an increase in the observed temperature and respiratory and pulse rates . The temperature increased to between 35–40 . 4°C with a mean of 38 . 18±0 . 7752°C ( normal range: 37 . 56–39 . 17°C ) and respiratory rates increased to 49-117/min with a mean of 85 . 25±13 . 91 breaths per minute ( normal range: 30–70 breaths per minute ) . The pulse rates ranged from 58–141 beats per minute with a mean of 98 . 78±18 . 64 beats per minute ( normal range: 120–180 beats per minute ) . A significant difference ( p = 0 . 045 ) was recorded in body temperature day 1 post treatment in the different treatment groups while no significant differences ( p>0 . 05 ) were recorded in the other vital signs between animals in the different treatment groups , ( Fig 4A–4C ) . The animals became withdrawn 48 hours after IVM treatment ( Fig 5 ) . Other clinical manifestations observed on this day were body rashes and itches ( Fig 6A ) , gland pain and gland tenderness , pinkish ears , swollen face , conjunctival haemorrhages , loss of appetite , muscle ache , slight discharge form the eyes and nostrils and diarrhea . The mean scores for rashes and itching observed in the various groups are shown in Fig 6B . Untreated animals did not show any clinical alterations . On the third day after the administration of ASA ( 1 tablet of 500mg 2 times a day for two days ) , the observed rashes and itches became absent or mild in the animals in this treatment arm whereas after the administration of PSE ( 20mg 2 times a day for two days ) , the observed rashes and itches remained moderate . The macroscopic changes seen included changes that were primarily those of a vasculo-pathologic nature ( e . g . haemorrhages , hyperaemia/injection , pleural effusions , etc . ) . Petechial haemorrhages were seen in the CNS , the lungs , the conjunctiva , the cardiac tissues , the peritoneum and the omentum . Adult worms were present in the deep subcutaneous tissues as well as the development of pleural effusions and enlarged and haemorrhagic lymph nodes , the skin showed excoriations and pachydermia on the limbs . These changes were seen in all groups given ivermectin , occurring at various levels , except in the untreated group ( which showed none of these changes ) . The majority ( 11/12 ) of baboons , all whom had been splenectomised as part of inducing the infection with L . loa , were seen to have regrown “new splenic tissue” . Most of these new tissues were variable in shape , located in the supra-renal area on the left or at least within 10 centimeters of this area; in two animals there were two independent such new organs in this anatomical location . These new tissues were most usually round ball-like and on sectioning were dark red in colour with a distinct fibrous capsule with some visible trabeculae . The different pathological changes seen are shown in Fig 7A–7F . Microfilariae were visible histologically in all animals . The brain and kidneys carried high numbers of mf in all treatment groups whereas lower numbers of mf were present in organs such as the diaphragm , pancreas and intestine in all other treatment groups except in the untreated infected group which recorded no mf in the diaphragm and pancreas . No significant differences ( p>0 . 05 ) were detected in the distribution of mf within the different organ tissues between any of the treatment groups including the kidneys ( p = 0 . 794 ) , except in the brain tissues , which had a significant more mf ( p = 0 . 04 ) in animals from different treatment groups . In untreated animals , mf was found in the tissues of all organs observed except in the diaphragm and pancreas . The highest numbers of mf counted in all observed tissues were in the brain ( 4 . 2 mf/25 mm2 and kidneys ( 5 . 4 mf/25 mm2 ) of these animals . The distribution of mf in the tissues of animals that served as controls can be seen in Fig 8 . In animals treated with IVM alone , mf were seen in all the observed tissues . In these animals , the mf increased greatly in most of the tissues . In the brain the increase to 13 . 8 mf/25 mm2 was 3 times; in the kidneys the decrease to 2 mf/25 mm2 in animals administered IVM alone was 3 times lower than that of animals infected but remained untreated and in the lymph nodes the increase to 4 mf/25 mm2 was 4 times-that of control animals . The distribution of mf in tissues from the different organs of animals that received IVM , can be seen in Fig 8 . In animals administered aspirin after ivermectin treatment , mf were also seen in the observed tissues . In these animals it was noticed that the mf seen in the tissues decreased greatly after ASA was administered . In the brain the decrease to 2 . 8 mf/25 mm2 ) was 5 times; kidneys the decrease to 8 mf/25 mm2 ) was 2 times; in the liver the decrease to 2 . 8 mf/25 mm2 ) was 2 times; in the lymph node the decrease to 1 . 8 mf/25 mm2 was 3 times -that of animals administered IVM alone . The distribution of mf in tissues from the different organs of animals that received IVM + ASA , can be seen in Fig 8 . Microfilariae were equally observed in the tissues of animals administered prednisone after IVM treatment . Here it was noticed that mf numbers increased greatly in the brain tissues . In the brain the increase in mf to 17 . 4 mf/25 mm2 was 1 . 3 times that of animals administered IVM alone . In the kidneys , liver and lungs decreases to 10 mf/25 mm2 , 3 . 8 mf/25 mm2 and 2 . 8 mf/25 mm2 , respectively , were noticed which were 0 . 9 , 1 . 5 and 0 . 6 times , respectively as compared to those of animals which were administered IVM alone . The distribution of mf in tissues from the different organs of animals that received IVM , can be seen in Fig 9B . The major cellular changes seen , and which were present in all the ivermectin treated animals , were: The lesions found in untreated animals and in the first 72 hours after treatment with ivermectin included eosinophilic accumulation in the tissues , accumulation of mf in the lymphoid tissues , fibrin deposition on the walls of cerebral blood vessels and mf degeneration and eosinophilic accumulation/degranulation in a lymph node . These changes can be seen in Fig 10A–10F . After 72 hr . following treatment , the observed lesions included blocked CNS vessels comprised of eosinophils , fibrin , macrophages and parasite debris , blocked CNS vessels with associated damage ( vacuolation of the parenchyma ) , intact microfilariae caught in a cellular intravascular mass in the CNS , degenerating mf in blood capillaries of the CNS and surrounded by fibrin , and an area of vascular and parenchymal damage in the CNS predominately filled with macrophages and eosinophils . The different lesions seen more than 72 hours post treatment with ivermectin can be seen in Fig 10A–10F . The presence and extent of histo-pathological changes seen in the different organs of the various treatment groups can be seen in Table 2 . The distribution of different lesions seen in the different tissues indicate that they were mild in untreated animals while in animals administered ivermectin alone , these lesions were moderate; the distribution of these lesions between treatment groups can be seen in Table 3 .
In this report the clinical spectrum and pathology related to the treatment of hypermicrofilaraemic Loa infected baboons with ivermectin is presented for the first time . After treatment with ivermectin , microfilaraemia decreased significantly in all treated animals with up to 98 . 4% reduction in one animal ( Bab 05 ) . This finding confirms the findings of earlier authors who had shown ivermectin to have a marked microfilaricidal effect on L . Loa . A study by Kamgno et al . [23] demonstrated that one month after a single dose with ivermectin , microfilarial loads fall to <20% of their initial values and this suppression of microfilaraemia can persist for at least a year [24] . The reduction of microfilariae numbers in peripheral blood could probably be due to the fact that ivermectin treatment induces the Loa microfilariae to flee from the blood circulation into other body fluids such as the pleural and peritoneal fluids as a result of vascular breaking or due to the death of these mfs . The drastic killing of circulatory mf could lead to the blockage of blood vessels in the different tissues and organs . When this happens , an inflammatory reaction sets in activating the fibrin pathway whereby fibrin clots are deposited in the tissues . The release of inflammatory cytokines such as interleukin ( IL ) -1 , IL-6 , IL-17 and tumor necrosis factor alpha ( TNF-α ) which have been incriminated as being part of pathological signal cascades in brain diseases [25] could have been released in these animals , although these cytokines were not measured . The formation of fibrin clots could result in the blockage of blood vessels in the different tissues and organs causing tissue anoxia and finally death . It is not clear as to the significant changes between the pre and post treatment haemoglobin levels . However , no animals were found to be clinically anaemic , and as the animals were fed on a rich protein-diet it is likely that this variation in parameter was not a major characteristic of ivermectin’s effect on haemoglobin . In all treated animals , there was an increase in the total white blood cell counts , eosinophils , and neutrophils . The increase in WBC counts is similar to what Ducorps et al . found out in their study [10] . This increase is probably due to an enhancement in the production of many more leukocytes to reduce the severity of Mazzotti reaction and to fight the infection in general . The increase in eosinophils in this study is similar to what Martin-Prevel et al . [26] noticed where treatment with ivermectin causes an increase in eosinophils . The global changes in eosinophil counts recorded after treatment with ivermectin in patients harbouring L . loa is similar to those observed in onchocerciasis patients after treatment with ivermectin or DEC [27 , 28] . This could probably be due to the preponderant role of eosinophils in the development of the Mazzotti reaction whose severity is associated with eosinophil sequestration and activation-degranulation [29 , 30] . The increase in neutrophils observed is probably because these cells are the hallmarks of acute inflammation given the fact that the migration and degranulation of eosinophil in the skin releases parasite specific antigens during cell death that induces a pro-inflammatory response . The general increase observed in SGPT , SGOT and γ-GT levels after ivermectin treatment could probably be due to hepatocellular damage or obstruction in bile flow cholestasis [31 , 32] . Generally , this drug has not been associated with acute or chronic liver injury , although Veit et al . , [33] realised that treatment with a single dose of ivermectin resulted in mild hepatotoxicity . Elevations in SGPT and SGOT in blood are increased in conditions in which cells are damaged or dead . It is also important to note that the elevations in serum liver enzymes can also be secondary to enzyme induction without hepatic pathology [34] . The elevations of creatinine post-treatment in control animals was not surprising because during the course of infection after inoculation of mfs into these animals , its’ levels were already high . This could probably be related to kidney dysfunction [35] as a malfunction in glomerular filtration results in the retention of substances including urea and creatinine . Generally , creatinine levels dropped in all animals that took ivermectin alone , or ivermectin together with other treatments . This decrease is probably due to the fact that the kidneys were more efficient in excreting the compound . The elevated serum glucose level observed may have resulted from increased mobilization of glucose for metabolism or may be due to reduced glucose uptake into cells caused by ivermectin [36] or a possible modulation of the capacity of the renal tubule by the drug to reabsorb glucose actively from the blood [36] . The decrease in potassium levels in untreated animals could probably be due to a reduced intake in their diet [37] or a decreased sensitivity of the nephron to aldosterone and other mineralocorticoids responsible for reabsorption and retention of electrolytes , respectively . However , the drop in potassium levels in animals administered ivermectin and prednisone is probably due to the fact that prednisone is known to cause more urination as more blood moves to the kidneys . It is probable that as more urine is sent out , more potassium ions are also lost along with it from the body in these animals or reduced intake in their diet [37] . The changes in calcium ions pre and post treatment could probably be related to variations in diet intake of this mineral [38] . Generally , all animals administered ivermectin became withdrawn 48 hours following treatment . This observation is in line with the results of earlier studies [12 , 13] which revealed that symptoms appeared within two days of ivermectin treatment in patients who were previously healthy . This withdrawal attitude could be due to stress that developed due to loss of coordination , reduced alertness that can progress to stupor or coma . A withdrawn attitude is a neurological manifestation that is related to blockage and reduced blood flow supplying nutrients and oxygen to the cells that could be due to the destruction or death of mf in the brain and other tissues . A possible scenario for the pathogenesis of the animal that died 5 hours after ivermectin treatment could be that at the time of ivermectin treatment , this animal harbored a Loa mf load ( 124 , 700mf/mL ) greatly exceeding the threshold associated with the risk of post-ivermectin Loa encephalopathy . The initial L . loa microfilarial load has been demonstrated to be the main risk factor for the development of serious reactions and an association has been found to exist between L . loa load and the occurrence of marked or serious reactions has been found to be significant above 8 , 000mf/mL with the association being close and stable [12 , 13] . In all animals treated with ivermectin , there was the development of body rashes and itches in otherwise healthy animals just one day after treatment . The development of rashes could probably be due to an enhancement in the production of many more leukocytes ( such as eosinophils and neutrophils ) to reduce the severity of Mazzotti reaction and to fight the infection in general; these cells being the hallmarks of acute inflammation . The presence of haemorrhages with evidence of L . loa mf seen in the brain , eyes and other organs of treated animals could probably be due to the fact that ivermectin results in rapid killing of Loa mf in these tissues . The obstruction of blood capillaries by these dead worms in these organs could lead to an ischemic reaction , increased pressure within the capillaries , rupture of the affected vessel ( s ) and haemorrhagic suffusion . This observed vasculopathy with evidence of L . loa as a likely etiology in the brain tissue therefore confirms the definite case of Loa encephalopathy as put forth by the group of independent experts who consulted for MDP immediately after the first cases of SAEs were noticed in the 1990s . Lymphadenopathy was present in all baboons that received ivermectin . Enlarged lymph nodes generally result from the accumulation of leucocytes within the lymph glands . Ivermectin , has been shown to cause an increase in the total white blood cell counts , number of eosinophils , and neutrophils [10] . Our results suggest that the pathogenesis of Loa encephalopathy is as a result of the death of mf in the blood vessels in the brain . This event causing inflammation to set in with the activation of the fibrin pathway with fibrin clots being deposited and these blocking the blood vessels in the brain and other tissues . This deposition of fibrin clots damages the walls of these blood vessels as a result of the release of inflammatory cytokines and these overall pathologic processes affecting the local tissues through tissue anoxia and other detrimental changes that eventually result in the permanent damage and in some cases the death of the host ( Fig 11 ) . For the first time , this study has provided information on the tissues and body fluids where mf could be found after ivermectin treatment of Loa heavily infected individuals . Generally , organs that normally receive large amounts of blood were noticed to be harbouring many more microfilariae than those less irrigated . Given that Loa loa is a blood dwelling filarial worm , its presence in these highly irrigated organs is to be expected . It is not known as to why aspirin and prednisone lead to a decrease or an increase in the number of tissue sections . The presence of microfilariae in other body fluids ( peritoneal and pericardial ) collected at autopsy could probably be due to the fact ivermectin treatment induces the Loa microfilariae to flee from the blood circulation . This is an important finding which indicates that in those humans who suffer from SAEs , the serious events noticed could be due to breathing difficulties due to the accumulation of fluid in the lungs or complications such as ascites which might cause difficulties in compressing the diaphragm leading to breathing difficulties . The discovery of the presence of these effusions could serve as an important step in managing the SAEs clinically as these fluids could be drained and an intravenous infusion administered to these patients to balance fluid and electrolytes . The primary lesions seen in the untreated infected animals were peri-portal accumulation of lymphocytes; as present in all of the animals . It is likely that in addition to the spleen , this organ and the peri-portal region is a common location for the degeneration of microfilariae . It is also possible that this occurs as a result of the initial removal of the spleen as part of the infection process , and the liver has become the default site for microfilarial destruction . It is clear that a major pathogenic event following ivermectin is a vasculopathy associated with the death and degeneration , with a cellular response , and all of this taking place intravascularly . Although there is on occasion clear damage and the resulting haemorrhaging of vessels , it is more likely that this vascular blockage results in acute parenchymal damage or dysfunction in the brain that results in the clinical condition . Generally , in this study we found out that animals administered aspirin showed milder symptoms compared to those that received prednisone . Aspirin is part of a group of medications called non-steroidal anti-inflammatory drugs ( NSAIDs ) , but differs from most other NSAIDs in the mechanism of action . It produces lipoxins , most of which are anti-inflammatory [39] . It is probable that the lipoxins produced by this drug aid in alleviating the symptoms observed in animals after treatment with ivermectin . The finding that prednisone aggravates the clinical manifestations observed corroborates the assertions made by Boussinesq et al . [40] that corticosteroids do not have any beneficial effect on the course of these clinical manifestations because they might lead to iatrogenic lethal complications . The influx of significant number of eosinophils into the tissues shortly after ivermectin treatment is likely to be a significant contributor to the pathogenesis of the condition or at least the clinical presentation . Release of toxic mediators from the activation of eosinophils would probably be an important factor in the pathogenesis of this condition .
The findings presented from this study have revealed that the ivermectin treatment of Loa hypermicrofilaraemic animals induces a drastic reduction in mf numbers with an 80% decrease observed in some animals . Hb was the only observed parameter that recorded a slight significant value pre and post treatment . Clinical manifestations observed after treatment are similar to what have been observed in humans such as rashes , itching , haemorrhagic conjunctiva , increased temperature , increased pulse and respiratory rates , diarrhoea , lymph node enlargement , pinkish ears , swollen face and restlessness . Macroscopic changes noticed were primarily those of a vasculo-pathologic nature ( e . g . haemorrhages ) seen in all groups given ivermectin , occurring at various levels , except in the untreated . Microscopically , the major cellular changes seen , and present in all the ivermectin treated animals , were: the presence of microfilariae of varying degrees of degeneration associated with either fibrin deposition , endothelial alteration ( cellular swelling ) including damage to the integrity of the blood vessel and the presence of extravascular erythrocytes ( haemorrhage ) , the increased presence of eosinophil leucocytes and other chronic inflammatory types in certain tissues and organs , sometimes in large numbers and local evidence of microfilarial death . Highly vascularised organs like the brain , heart , lungs , and kidneys were observed to have more microfilariae in tissue sections . The number of mf seen in the brain and kidneys of animals administered IVM only , tripled that of control animals . Co-administration of IVM + PSE caused an increase in mf in the brain and kidneys while the reverse was noticed with the co-administration of IVM+ASA . The ivermectin treatment of Loa hyper-microfilaraemic baboons has revealed that the formation of fibrin clots and the presence of pleural and peritoneal effusions could probably be related to the development of the pathological conditions noticed in those Loa hypermicrofilaraemic humans who were administered ivermectin .
|
The pathogenesis of the serious adverse reactions that occur in patients carrying high loads ( usually > 30 , 000mf/mL ) of circulating Loa loa parasites and treated with ivermectin has to date not been clearly defined at the tissue level . These reactions , which can result in the death of many of those affects and cause permanent mental damage in others , inflicts considerable disruption to the important mass treatment programs to control and eliminate filariae where ivermectin is used . The recently described baboon model of hyper-microfilariaemia has been used here to study the pathological changes occurring after ivermectin treatment; the clinico-pathological presentation following treatment in baboons appears to be closely related to those seen in humans . The tissue changes seen suggest that the primary pathological event following ivermectin treatment is a vasculopathy associated with intra-vascular microfilarial destruction within smaller blood vessels that damages them and leads to changes in the tissues supplied by these vessels . The effects of cortisone and aspirin on these changes are also addressed in this study .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion",
"Conclusions"
] |
[
"blood",
"cells",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"immunology",
"vertebrates",
"fibrin",
"animals",
"mammals",
"primates",
"lymph",
"nodes",
"signs",
"and",
"symptoms",
"lymphatic",
"system",
"kidneys",
"old",
"world",
"monkeys",
"white",
"blood",
"cells",
"monkeys",
"baboons",
"animal",
"cells",
"proteins",
"loa",
"loa",
"hematology",
"biochemistry",
"eosinophils",
"diagnostic",
"medicine",
"cell",
"biology",
"anatomy",
"hemorrhage",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"renal",
"system",
"vascular",
"medicine",
"amniotes",
"organisms"
] |
2017
|
Ivermectin treatment of Loa loa hyper-microfilaraemic baboons (Papio anubis): Assessment of microfilarial load reduction, haematological and biochemical parameters and histopathological changes following treatment
|
Sphingolipids have essential roles as structural components of cell membranes and in cell signalling , and disruption of their metabolism causes several diseases , with diverse neurological , psychiatric , and metabolic consequences . Increasingly , variants within a few of the genes that encode enzymes involved in sphingolipid metabolism are being associated with complex disease phenotypes . Direct experimental evidence supports a role of specific sphingolipid species in several common complex chronic disease processes including atherosclerotic plaque formation , myocardial infarction ( MI ) , cardiomyopathy , pancreatic β-cell failure , insulin resistance , and type 2 diabetes mellitus . Therefore , sphingolipids represent novel and important intermediate phenotypes for genetic analysis , yet little is known about the major genetic variants that influence their circulating levels in the general population . We performed a genome-wide association study ( GWAS ) between 318 , 237 single-nucleotide polymorphisms ( SNPs ) and levels of circulating sphingomyelin ( SM ) , dihydrosphingomyelin ( Dih-SM ) , ceramide ( Cer ) , and glucosylceramide ( GluCer ) single lipid species ( 33 traits ) ; and 43 matched metabolite ratios measured in 4 , 400 subjects from five diverse European populations . Associated variants ( 32 ) in five genomic regions were identified with genome-wide significant corrected p-values ranging down to 9 . 08×10−66 . The strongest associations were observed in or near 7 genes functionally involved in ceramide biosynthesis and trafficking: SPTLC3 , LASS4 , SGPP1 , ATP10D , and FADS1–3 . Variants in 3 loci ( ATP10D , FADS3 , and SPTLC3 ) associate with MI in a series of three German MI studies . An additional 70 variants across 23 candidate genes involved in sphingolipid-metabolizing pathways also demonstrate association ( p = 10−4 or less ) . Circulating concentrations of several key components in sphingolipid metabolism are thus under strong genetic control , and variants in these loci can be tested for a role in the development of common cardiovascular , metabolic , neurological , and psychiatric diseases .
Sphingolipids are essential components of plasma membranes and endosomes and are believed to play critical roles in cell surface protection , protein and lipid transport and sorting , and cellular signalling cascades . They are known to have roles in both health and disease [1] , [2] . Several rare monogenic diseases associated with sphingolipid biosynthesis and turnover have been identified such as metachromatic leukodystrophy and GM1- and GM2-gangliosidosis , Niemann-Pick , Gaucher , Krabbe , Fabry , Farber , Tay-Sachs and Sandhoff diseases [3] . Defective biosynthesis due to mutations in genes involved in sphingolipid metabolism ( e . g . serine palmitoyl transferase ( SPTLC1 ) [4]; ceroid-lipofuscinosis , neuronal 8 ( CLN8 ) [5]; and ceramide synthase ( LASS1 ) [6] ) can also lead to disease . Moreover , natural fungal inhibitors of ceramide synthase can result in a broad spectrum of effects including equine leucoencephalomalacia , porcine pulmonary oedema syndrome and liver cancer in rats [7] , demonstrating the wide range of processes that include cell proliferation , differentiation and apoptosis underpinned by sphingolipid metabolism . Identifying common genetic variants that influence the balance between individual sphingolipid concentrations represents an important step towards understanding the contribution of sphingolipids to common human disease . To achieve this goal , we conducted a genome-wide association study ( GWAS ) on plasma levels of 33 major sphingolipid species ( 24 sphingomyelins and 9 ceramides ) in five European populations , both within and across populations . The traits were analysed by individual species ( sphingomyelins ( SM ) , dihydrosphingomyelins ( Dih SM ) , ceramides ( Cer ) and glucosylceramides ( GluCer ) ) or aggregated into groups of species with similar characteristics ( e . g . unsaturated ceramides ) , and expressed as absolute concentrations or as molar percentages within sphingolipid classes ( mol% ) . In addition we examined 43 matched metabolite ratios between the traits as a surrogate for enzyme activity [8] in separate clusters designed to examine sphingolipid metabolism ( 11 ratios ) , desaturation ( 16 ratios ) and elongation ( 16 ratios ) . All traits displayed substantial heritabilities in that much of the observed variation in sphingolipid levels could be attributed to genetic variation among individuals in each population .
The GWAS for single species and matched metabolite ratios revealed a total of 32 SNPs in five distinct loci reaching genome-wide significance ( p values ranging down to 9 . 08×10−66 ) ( Table 1 , Figure 1 and Figure 2 , and Table S1 and Table S3 ) . The direction and magnitude of the observed effect sizes for the 22 variants identified in the analysis of single species are summarized in Table 1 with full details in Table S1 . For three of the regions ( chromosomal regions 4p12 , 14q23 . 2 and 19p13 . 2 ) , p values reached genome-wide significance in the largest cohort ( South Tyrol ) , and the effect was replicated in the other populations . For two additional loci ( 11q12 . 3 and 20p12 . 1 ) , signals bordered on genome-wide significance in South Tyrol alone , were consistent between all 5 populations and reached genome-wide significance in the meta-analysis . In the single species analysis , the strongest associations for three of the loci ( 11q12 . 3 , 14q23 . 2 and 19p13 . 2 ) were found with sphingomyelins and dihydrosphingomyelins . The 4p12 locus showed the strongest association with serum glucosylceramides and the 20p12 . 1 locus showed the strongest association with serum ceramide concentrations . Table S2 shows the p-values for the individual SNPs when included in a multiple regression model , and the fraction of single sphingolipid variance explained by sex , age and all SNPs in the model together . Taken together , the SNPs explain up to 10 . 1% of the population variation in each trait . Ratios of matched ( substrate/product ) pairs have been shown to reduce variation in the dataset and increase power of association several orders of magnitude [8] . Analysis of 43 matched metabolite ratios ( Table S3 ) indeed increased power of association up to 10 orders of magnitude on some of the 22 variants above , and revealed an additional 10 SNPs over the same 7 genes reaching statistical significance ( see Table S3 ) . Surprisingly no signals from new genes reached genome-wide significance , highlighting the fact that across the 5 cohorts analysed here , the 7 genes identified are the major genes associated with circulating sphingolipid concentrations . Among the 32 significant individual SNPs ( Table S4 ) variants in LASS4 explain up to 7 . 5% of the variance in some ratios ( i . e . in SM16:0/SM18:0 ) , SGPP1 variants explain up to 12 . 7% of the variance ( i . e in SM14:0/SM16:0 ) , FADS1–3 variants explain up to 3 . 5% of the variance ( e . g . in SM16:0/SM16:1 ) , SPTLC3 variants explain up to 4 . 9% of the variance ( e . g . in SM14:0/SM16:0 and SM24:0/Cer24:0 ) , and ATP10D variants up to 4 . 2% of GluCer/Cer variance . Combined effects of several genes ( i . e . SPTLC3 and SGPP1 ) explains up to 14 . 2% of the variance in medium chain SM ratios ( SM14:0/SM16:0 ) and , in combination with LASS4 , up to 11 . 2% of the variance in long-chain sphingomyelin ratios ( SM22:0/SM24:0 ) . All SNPs within the associated chromosomal regions are located within or are in linkage disequilibrium ( LD ) with genes that encode enzymes involved in sphingolipid biosynthesis or intracellular transport ( Figure 2 ) . The ATPase , class IV , type 10D ( ATP10D ) gene , located at chromosome 4p12 , encodes a putative serine-phospholipid ( phosphatidylserine , ceramide ) translocase [9] . Three SNPs at this locus showed genome-wide significant associations with glucosylceramides ( C16:0 , C24:1 ) ( Table 1 , Table S1 ) , with an additional five variants revealed in the ratio analysis ( Table S3 ) . SNP rs10938494 gave the strongest association in the single species analysis ( p-values of 1 . 68×10−9 in South Tyrol and 8 . 03×10−19 in the joint analysis ) , and was among the strongest association in the ratio analysis ( p = 3 . 04×10−16 ) along with rs2351791 ( p = 6 . 58×10−17 ) . Three fatty-acid desaturase genes ( FADS1 , 2 and 3 ) are located adjacent to one another in a cluster at the 11q12 . 3 locus . The FADS1–3 genes encode enzymes that regulate the desaturation of fatty acids by the introduction of double bonds between defined carbons of the fatty acyl chain . Seven SNPs at this locus , distributed in and around the three genes , reached statistical significance in the single species analysis for sphingomyelin 16∶1 levels in the joint analysis , with p-values ranging from 2 . 99×10−11 ( rs174449 , close to FADS3 ) to 6 . 60×10−13 ( rs1000778 , in FADS3 ) ( Table 1 ) . The ratio analysis revealed an additional SNP at this locus within the FADS3 gene ( rs174450 , Table S3 ) , and improved association results for other SNPs several orders of magnitude ( e . g . rs1000778 p = 1 . 29×10−15 ) . Fatty acids are built into ceramides by the ceramide synthases ( e . g . LASS4 ) . Unsaturated ceramides can be synthesized exclusively by the introduction of unsaturated fatty acids into the sphingosine/sphinganine chain . The pivotal role of FADS1–3 in the synthesis of unsaturated ceramides is confirmed by the strong associations of SNPs in this cluster to the mono-unsaturated sphingomyelins 16∶1 , 18∶1 and 20∶1 , which are the end-products of the ceramide biosynthesis pathway ( Table 1 , Table S1 ) , and the ratios between these and their respective unsaturated precursors ( Table S3 ) . Previous studies of sphingolipid metabolites and poly-unsaturated fatty acids ( PUFA ) have demonstrated associations to SNPs , including rs174537 , over the FADS1 and FADS2 genes in several populations [8] , [10] , [11] . The sphingosine-1-phosphate phosphohydrolase 1 gene ( SGPP1 ) at the 14q23 . 2 locus belongs to the super-family of lipid phosphatases that catalyze the generation of sphingosine and , together with irreversible cleavage by sphingosine-1-phosphate ( S1P ) -lyase , strongly influences the pathway of S1P to ceramide ( Figure 3 ) . Six SNPs in and around this gene demonstrate the most significant associations with circulating sphingomyelin C14–C16/C22–C24 and dihydrosphingomyelin concentrations ( Table 1 ) in the single species analysis , with a further two SNPs revealed in the ratio analysis . SNP rs7157785 showed the strongest association with sphingomyelin 14∶0 relative content ( molar percentage: mol% ) with genome-wide significant p-values in all five populations , particularly in the South Tyrol population ( p = 2 . 53×10−28 ) and joint analysis ( p = 9 . 08×10−66 ) , and demonstrated the most significant association in the ratio analysis . Enhanced SGPP1 activity leads to elevated ceramide levels by shifting the stochiometric balance of SGPP1/S1P-lyase towards sphingosine and ceramide production . Five SNPs at the 19p13 . 2 locus showed some of the strongest associations with sphingolipids and all lie within LASS4 , the gene encoding LAG1 longevity assurance homologue 4 . In the single species analysis SNP rs7258249 showed the highest genome-wide significant association with sphingomyelin 18∶0 mol% ( South Tyrol p = 1 . 04×10−15 and joint analysis p = 2 . 28×10−27 ) . Several LASS4 SNPs showed statistically significant association with the sphingomyelin species C18 to C20 and with ceramide C20∶0 ( Table 1 and Table S1 ) . In the ratio analysis , however , associations strengthened by several orders of magnitude ( p value ) over those with these SNPs , with rs1466448 demonstrating the most statistically significant association ( p = 4 . 05×10−35 ) . LASS family members , six of which have been identified in mammals ( LASS1–6 ) , are de novo ceramide synthases ( CerS ) that synthesize dihydroceramide from sphinganine and fatty acid ( Figure 3 ) . Moreover , LASS enzymes catalyze the re-synthesis of ceramide and phytoceramide from sphingosine and phytosphingosine respectively , which are cleavage products of alkaline ceramidase activity in endoplasmic reticulum ( ER ) membranes . The 20p12 . 1 locus contains the serine palmitoyltransferase long chain base subunit 3 gene ( SPTLC3 ) encoding a functional subunit of the SPT enzyme-complex that catalyzes the first and rate-limiting step of de novo sphingolipid synthesis . One SNP ( rs680379 ) demonstrated association for unsaturated ceramide in the South Tyrol population alone ( p = 1 . 77×10−07 ) and was genome-wide significant in the joint analysis ( p = 8 . 24×10−15 ) . Significant association was observed also with C16 to C24 ceramides and the sphingomyelins 16∶1 and 17∶0 ( Table 1 and Table S1 ) . The ratio analysis strengthened association at this variant ( p = 3 . 3×10−20 for the metabolite ratio SM24:0/Cer24:0 ) and revealed two further significant variants at this locus ( rs3848751 and rs6078866 , Table S3 ) . As matched metabolite ratios can serve as a proxy for enzyme activity [8] , in a complementary candidate gene approach , we investigated association signals in our combined single species and ratio datasets at 624 SNPs within or near 40 genes that encode enzymes involved in sphingolipid metabolism , in order to identify the most promising variants within these genes for further analysis . Of these , a total of 70 variants in or near 23 of the genes demonstrate association p values of 10−4 or less ( Table S5 ) . Sex and age adjusted single sphingolipids species displayed strong phenotypic correlations with circulating plasma lipoproteins especially with total cholesterol or LDL-cholesterol ( Table S6 , e . g . between the sum of saturated sphingomyelin species and total cholesterol: 0 . 788/0 . 717/0 . 794/0 . 733/0 . 773 in respectively NPHS/ERF/SOUTH TYROL/CROATIA/ORKNEY; or SM16:1 and total cholesterol 0 . 737/0 . 631/0 . 671/0 . 6/0 . 638 ) . This is in agreement with recent lipid profiling of lipoprotein fractions , showing higher proportions of sphingomyelin and ceramides in the LDL fraction [12] . However , among the GWAS hits uncovered in this analysis , only the FADS1–3 cluster overlaps with those reported in large meta-analysis of circulating serum lipoproteins levels ( strongest with total and LDL-cholesterol levels ) [13] . Several of the variants reported here display suggestive associations with classical lipids in the EUROSPAN cohorts ( Table S7 ) . All eight SNPs in the FADS1–3 cluster associate with HDL-cholesterol levels ( age-sex adjusted p values between 0 . 06 and 0 . 0041 ) similar to previous observations [8] . Interestingly , the sex-specific age-adjusted results show that these associations seem driven by the association found in males ( lowest p = 0 . 0037 at rs174546 ) . Association with HDL-cholesterol in males is also seen with SNPs in ATP10D ( rs2351791 , p = 0 . 01 ) and SPTLC3 ( rs3848751 , p = 0 . 0047 ) . SNPs at ATP10D also associate with LDL-cholesterol , albeit weakly in the total population ( rs469463 , p = 0 . 034 ) . In the males only , variants at LASS4 ( rs28133 , p = 0 . 043 ) and SPTLC3 ( rs3848751 , p = 0 . 022 and rs6078866 , p = 0 . 02 ) also associate weakly with LDL-cholesterol levels . Five variants in FADS1–3 and two in ATP10D associate with triglyceride levels , with lower p values in males than in the whole group ( p values from 0 . 017 to 0 . 009 in FADS1–3 and 0 . 0071 for rs17462424 in ATP10D ) . Association of FADS variants with triglyceride levels has also been observed in other populations [8] . As previously highlighted [8] , the p values for association with the sphingolipids species were orders of magnitude stronger than with these classical lipids . Given the reported associations to classical lipids and cardiovascular disease with variants at the FADS1–3 locus [10] , [13] , [14] , and the evidence from functional studies of a role for sphingolipids in atherosclerotic plaque formation and lipotoxic cardiomyopathy [15] , we looked in silico in a series of three age- and sex-adjusted GWAS datasets of German myocardial infarction ( MI ) case-control studies ( Ger MIFS I [16] Ger MIFS II [17] and Ger MIFS III ( KORA ) , unpublished ) for evidence of association with the major variants associating with sphingolipid concentrations . Variants within three of the genes ( ATP10D , FADS3 and SPTLC3 ) associate with MI in one or more of the studies ( Table 2 ) . The protective odds ratios observed for variants in ATP10D and SPTLC3 are on alleles correlating positively with higher metabolite/lower ceramide ratios ( i . e . GluCer/Cer and SM/Cer ) , in support of evidence that increased enzyme/transporter activity that lowers ceramide levels might alleviate the pro-apoptotic effects seen with higher ceramide levels in cardiomyocytes [18] . As previously hypothesised , carriers of FADS variants that are associated with higher desaturase activity may be prone to a proinflammatory response favoring atherosclerotic vascular damage [14] .
Direct experimental evidence indicates a role for sphingolipids in several common complex chronic disease processes including atherosclerotic plaque formation , myocardial infarction ( MI ) , cardiomyopathy , pancreatic beta cell failure , insulin resistance and type 2 diabetes mellitus ( T2D ) [15] . Until now , the genetic variants that influence circulating sphingolipid concentrations in the general population have been characterized in relatively small cohorts [8] . Here we identified genetic variation with a significant effect on the biosynthesis , metabolism or intracellular trafficking of some of the major sphingolipids species in a large diverse group of European population samples . The SNPs showing association with circulating sphingolipids explain up to 10 . 1% of the population variation in each trait and 14 . 2% of some matched ratios ( Tables S2 and Table S4 ) . Four of the five loci identified contain genes encoding proteins that are either responsible for de novo ceramide synthesis or for ceramide re-synthesis from sphingosine/sphinganine-phosphates or both ( SPTLC3 , LASS4 , FADS1–3 and SGPP1 ) . Increases in all of these enzymatic activities are predicted to elevate the “ceramide-pool” . The associations are observed not only with ceramides , but also with sphingomyelins , indicating that a considerable proportion of ceramide is converted into the large and more stable “sphingomyelin-pool” . None of the genes involved in ceramide degradation or ceramide-related signaling is genome-wide significantly associated with the traits analyzed , indicating the primary role of genes related to ceramide production in the genetic control of ceramide levels . Of these four loci , the FADS1–3 gene cluster has been the most frequently to be reported linked with disease in recent literature . Variants within in this region have been associated with cardiovascular disease and classic lipid risk factors such as cholesterol levels [10] , [13] , [14] . Reported variants demonstrating association in these reports ( rs174547 , rs174570 , rs174537 and rs174546 ) are within the FADS1 and FADS2 genes , but expression studies indicate complex regulation in this region , with the FADS1 SNP rs174547 showing correlation with expression of both FADS1 and FADS3 genes [19] , while the FADS1 SNP rs174546 correlates with FADS1 but not FADS2 expression [10] . Our strongest associations with both sphingolipid levels and MI are in or nearest the FADS3 gene , with variants showing less marked association with cholesterol levels than that observed with variants over FADS1 and FADS2 genes ( Table S7 ) . It is known that sphingomyelin and ceramides can modulate the atherogenic potential of LDL [20] . Further functional studies will be necessary to determine whether the active mechanism is through FADS3 alone , or in concert with FADS1 , FADS2 or both . Neurological phenotypes associated with FADS2 include attention-deficit/hyperactivity disorder [21] and the moderation of breastfeeding effects on IQ [22] . Little is published regarding disease association with variants at the other four major loci described here . However , a reported association between expression levels of SGPP1 with Schizophrenia [23] along with changes in SPTLC2 ( with variants identified in our candidate SNP search –Table S4 ) and ASAH1 , highlights the importance of testing variants in these genes with multiple neurological and psychiatric diseases . Additional neurological associations with candidate genes listed in Table S4 include SGPL1 in Alzheimer's disease [24] and GBA with Parkinson's disease and dementia with Lewy bodies [25] , [26] . The wider possible involvement of genes within pathways of ceramide metabolism in Lewy body disease has also been recently reviewed [27] . The fifth locus contains ATP10D , a cation transport ATPase ( P-type ) type IV subfamily member . The type IV subfamily is thought to be an important regulator of intracellular serine-phospholipid trafficking however the exact function or transport specificity of ATP10D has not yet been described [9] . A novel functional finding of this study is the specificity of the association of ATP10D SNPs to glucosylceramides ( among the species tested so far ) , which provides the first evidence for the functional involvement of ATP10D in intracellular transport of specific species of ceramide ( Figure 3 ) . Impaired function of ATP10D may therefore lead to enhanced exposure of ceramide to glucosyltransferases , forming higher concentrations of glycosylceramides that are released into the plasma compartment or may elevate serum glucosylceramide concentrations by impaired transport of glycosylceramide to the trans Golgi network . Mutations of ATP10D ( C57BL/6J ( B6 ) ) in mice result in low HDL concentrations and these mice develop severe obesity , hyperglycaemia and hyperinsulinaemia when fed on a high-fat diet [28] . Based on the mouse model , increased circulating glucosylceramides in connection with ATP10D function would be one plausible mechanism of contributing to weight gain and early insulin resistance . From the novel association of SNPs in ATP10D to MI ( Table 2 ) seen in German studies , further investigation of the specific role of glucosylceramides in MI and other cardiovascular diseases is warranted . Thus , sphingolipids play a role in pathological processes leading to common complex diseases , and identification of genetic variants that influence the balance between individual sphingolipid species is an important first step into dissecting out the genetic components in such processes . Associations between the SNPs identified in this study , some of which have a strong effect on the circulating plasma levels , and complex metabolic , cardiovascular , inflammatory and neurological diseases in which a role for a sphingolipid-dependent mechanism is implicated can now be investigated . Modulation of sphingolipids in vivo has demonstrated that this may be a successful preventative strategy for diseases in which sphingolipids play a role , lending hope that , once such disease contributions are identified , successful therapeutic regimes may subsequently be identified .
All studies were approved by the appropriate Research Ethics Committees . The Northern Swedish Population Health Study ( NSPHS ) was approved by the local ethics committee at the University of Uppsala ( Regionala Etikprövningsnämnden , Uppsala ) . The ORCADES study was approved by the NHS Orkney Research Ethics Committee and the North of Scotland REC . The Vis study was approved by the ethics committee of the medical faculty in Zagreb and the Multi-Centre Research Ethics Committee for Scotland . The ERF study was approved by the Erasmus institutional medical-ethics committee in Rotterdam , The Netherlands . The MICROS study was approved by the ethical committee of the Autonomous Province of Bolzano . For the German MI studies ( GerMIFS-I , -II and –III ( KORA ) , local ethics committees approved the studies and written informed conset obtained as published previously . The ERF study is a family-based study which includes over 3000 participants descending from 22 couples living in the Rucphen region in the 19th century . All descendants were invited to visit the clinical research center in the region where they were examined in person and where blood was drawn ( fasting ) . Height and weight were measured for each participant . All participants filled out questionnaire on risk factors , including smoking . The 800 participants included in the lipidomics studies consisted of the first series of participants . The MICROS study is part of the genomic health care program ‘GenNova’ and was carried out in three villages of the Val Venosta on the populations of Stelvio , Vallelunga and Martello . This study was an extensive survey carried out in South Tyrol ( Italy ) in the period 2001–2003 . An extensive description of the study is available elsewhere [29] . Briefly , study participants were volunteers from three isolated villages located in the Italian Alps , in a German-speaking region bordering with Austria and Switzerland . Due to geographical , historical and political reasons , the entire region experienced a prolonged period of isolation from surrounding populations . Information on the health status of participants was collected through a standardized questionnaire . Laboratory data were obtained from standard blood analyses . Genotyping was performed on just under 1400 participants with 1334 available for analysis after data cleaning . All participants were included in the lipidomics studies . The Swedish samples are part of the Northern Swedish Population Health Study ( NSPHS ) representing a family-based population study including a comprehensive health investigation and collection of data on family structure , lifestyle , diet , medical history and samples for laboratory analyses . Samples were collected from the northern part of the Swedish mountain region ( County of Norrbotten , Parish of Karesuando ) . Historic population accounts show that there has been little immigration or other dramatic population changes in this area during the last 200 years . The Orkney Complex Disease Study ( ORCADES ) is an ongoing family-based , cross-sectional study in the isolated Scottish archipelago of Orkney . Genetic diversity in this population is decreased compared to Mainland Scotland , consistent with the high levels of endogamy historically . Data for participants aged 18 to 100 years , from a subgroup of ten islands , were used for this analysis . Fasting blood samples were collected and over 200 health-related phenotypes and environmental exposures were measured in each individual . All participants gave informed consent and the study was approved by Research Ethics Committees in Orkney and Aberdeen . The Vis study includes a 986 unselected Croatians , aged 18–93 years , who were recruited into the study during 2003 and 2004 from the villages of Vis and Komiza on the Dalmatian island of Vis [30] , [31] . The settlements on Vis island have unique population histories and have preserved their isolation from other villages and from the outside world for centuries . Participants were phenotyped for 450 disease-related quantitative traits . Biochemical and physiological measurements were performed , detailed genealogies reconstructed , questionnaire of lifestyle and environmental exposures collected , and blood samples and lymphocytes extracted and stored for further analyses . Samples in all studies were taken in the fasting state . Lipids were quantified by electrospray ionization tandem mass spectrometry ( ESI-MS/MS ) in positive ion mode as described previously [32] , [33] . EDTA plasma ( serum for South Tyrol ) samples were quantified upon lipid extraction by direct flow injection analysis using the analytical setup described by Liebisch et al . [33] . A precursor ion scan of m/z 184 specific for phosphocholine containing lipids was used for phosphatidylcholine ( PC ) and sphingomyelin ( SM ) [33] . Ceramide and hexosylceramide were analyzed using a fragment ion of m/z 264 [32] . For each lipid class two non-naturally occurring internal standards were added and quantification was achieved by calibration lines generated by addition of naturally occurring lipid species to plasma . Deisotoping and data analysis for all lipid classes was performed by self programmed Excel Macros according to the principles described previously [33] . Nomenclature of sphingomyelin species is based on the assumption that d18∶1 ( dihydroxy 18∶1 sphingosine ) is the main base of plasma sphingomyelin species , where the first number refers to the number of carbon atoms in the chain and the second number to the number of double bonds in the chain . DNA samples were genotyped according to the manufacturer's instructions on Illumina Infinium HumanHap300v2 ( except for samples from Vis for which version 1 was used ) or HumanCNV370v1 SNP bead microarrays . Four populations have 318 , 237 SNP markers in common that are distributed across the human genome , with Vis samples having 311 , 398 SNPs in common with the other populations . Samples with a call rate below 97% were excluded from the analysis . Sphingolipid measurements were available for analysis following quality control assessment for 4110 study participants . Genome-wide association analysis was performed using the GenABEL package in R [34] . A score test was used to test for association between the age- and sex-adjusted residuals of sphingolipid traits ( both as absolute concentrations and as relative content of the total sphingolipid pool: mol% ) and SNP genotypes using an additive model . The Genomic Control procedure [35] was used to account for under-estimation of the standard errors of effects , which occurs because of pedigree structure present in the data [36] . For the most interesting results and the species ratios , we re-analysed the data using “mmscore” function , a score test for family-based association [37] , as implemented in GenABEL . The relationship matrix used in analysis was estimated using genomic data with “ibs” ( option weight = “freq” ) function of GenABEL . This analysis , accounting for pedigree structure in an exact manner , allowed for unbiased estimation of the effects of the genetic variants ( adjusted for age and sex ) . The results from all cohorts were combined into a fixed-effects meta-analysis with reciprocal weighting on standard errors of the effect-size estimates , using MetABEL ( http://mga . bionet . nsc . ru/~yurii/ABEL/ ) . Thresholds for genome wide significance were set at a p value of less than 1 . 57×10−7 ( 0 . 05/318 , 237 SNPs ) for the individual populations . For the overall meta-analysis we chose to use the conservative threshold of 7 . 2×10−8 [38] . Since many of the traits tested and especially the ratios demonstrate high degrees of correlation , introducing a suitable statistical correction the multiple testing of the 76 correlated traits would be complex . Since Bonferroni correction ( unsuitable in this instance ) would lower thresholds to values between p = 10−9 to 10−10 , and since all five genomic regions have variants with p values <10−10 , we report the age-sex corrected p values alone . The threshold for replication of significant results from one population in other cohorts was set at a p-value less than 0 . 05 divided by the number of SNPs tested . All significant variants reported are in Hardy-Weinberg Equilibrium , and effect directions are consistent across all five populations .
|
Although several rare monogenic diseases are caused by defects in enzymes involved in sphingolipid biosynthesis and metabolism , little is known about the major variants that control the circulating levels of these important bioactive molecules . As well as being essential components of plasma membranes and endosomes , sphingolipids play critical roles in cell surface protection , protein and lipid transport and sorting , and cellular signalling cascades . Experimental evidence supports a role for sphingolipids in several common complex chronic metabolic , cardiovascular , or neurological disease processes . Therefore , sphingolipids represent novel and important intermediate phenotypes for genetic analysis , and discovering the genetic variants that influence their circulating concentrations is an important step towards understanding how the genetic control of sphingolipids might contribute to common human disease . We have identified 32 variants in 7 genes that have a strong effect on the circulating plasma levels of 33 distinct sphingolipids , and 43 matched metabolite ratios . In a series of 3 German MI studies , we see association with MI for variants in 3 of the genes tested . Further cardiovascular , metabolic , neurological , and psychiatric disease associations can be tested with the variants described here , which may identify additional disease risk and potentially useful therapeutic targets .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroscience/motor",
"systems",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration",
"cardiovascular",
"disorders/myocardial",
"infarction",
"genetics",
"and",
"genomics/medical",
"genetics",
"genetics",
"and",
"genomics/population",
"genetics"
] |
2009
|
Genetic Determinants of Circulating Sphingolipid Concentrations in European Populations
|
Scd6 protein family members are evolutionarily conserved components of translationally silent mRNA granules . Yeast Scd6 interacts with Dcp2 and Dhh1 , respectively a subunit and a regulator of the mRNA decapping enzyme , and also associates with translation initiation factor eIF4G to inhibit translation in cell extracts . However , the role of Scd6 in mRNA turnover and translational repression in vivo is unclear . We demonstrate that tethering Scd6 to a GFP reporter mRNA reduces mRNA abundance via Dcp2 and suppresses reporter mRNA translation via Dhh1 . Thus , in a dcp2Δ mutant , tethered Scd6 reduces GFP protein expression with little effect on mRNA abundance , whereas tethered Scd6 has no impact on GFP protein or mRNA expression in a dcp2Δ dhh1Δ double mutant . The conserved LSm domain of Scd6 is required for translational repression and mRNA turnover by tethered Scd6 . Both functions are enhanced in a ccr4Δ mutant , suggesting that the deadenylase function of Ccr4-Not complex interferes with a more efficient repression pathway enlisted by Scd6 . Ribosome profiling and RNA-Seq analysis of scd6Δ and dhh1Δ mutants suggests that Scd6 cooperates with Dhh1 in translational repression and turnover of particular native mRNAs , with both processes dependent on Dcp2 . Our results suggest that Scd6 can ( i ) recruit Dhh1 to confer translational repression and ( ii ) activate mRNA decapping by Dcp2 with attendant degradation of specific mRNAs in vivo , in a manner dependent on the Scd6 LSm domain and modulated by Ccr4 .
After being transcribed and processed in the nucleus , and exported to the cytoplasm , mRNAs can either engage with the translational machinery for protein synthesis , undergo storage in a translationally silent state , or be targeted for degradation . Cellular mRNAs can alternate between these processes , and translation , storage , and decay influence each other in multiple ways to regulate gene expression [1] . In general , mRNAs selected for translation are thought to establish interactions of their 5’-cap and 3’-poly ( A ) tail appendages with proteins from the translational machinery . mRNAs are activated for translation by binding to the mRNA cap of eukaryotic initiation factor eIF4F ( comprised of cap-binding protein eIF4E , scaffolding protein eIF4G , and helicase eIF4A ) and association of poly ( A ) -binding protein ( PABP ) with the poly ( A ) tail; and interactions between eIF4G and PABP can form a “closed-loop” mRNP competent for initiation . Further interactions between eIF4G and other translation initiation factors associated with the 43S pre-initiation complex ( PIC ) pre-assembled on the small ( 40S ) ribosomal subunit serve to recruit mRNA and form a 48S PIC competent for subsequent mRNA scanning and start codon selection ( reviewed in [2] ) . The mRNA decay machinery can compete with the translation initiation machinery for access to cap and poly ( A ) tail of the transcript [1] . Degradation of mRNA is generally initiated by removal of the poly ( A ) tail through sequential deadenylation reactions by the Pan2/Pan3 and Ccr4-Not complexes , followed by loss of associated PABP [3] . Turnover can proceed in the 3’ to 5’ direction via the cytoplasmic exosome , or in the 5’ to 3’ direction via removal of the cap by the Dcp2/Dcp1 enzyme complex and exonucleolytic digestion by Xrn1 [4] . Decapping is a highly regulated step that irreversibly commits mRNAs for complete digestion [5] . One regulatory mechanism is thought to include formation of complexes comprised of Dcp1/2 , Xrn1 , and distinct sets of Dcp2 interactors that modulate both mRNA substrate recruitment or catalysis by Dcp2 [4 , 6] . These decapping activators include DEAD-box helicase Dhh1 , Pat1 , Edc3 , the Lsm1-7 complex , and Scd6 [4] . Decapping activators can function by distinct mechanisms: ( i ) Scd6 , Dhh1 , and Pat1 can inhibit translation initiation by blocking formation of a 48S PIC in vitro , which in turn favors mRNA decapping [7]; ( ii ) Pat1 and Edc3 can directly bind Dcp2 and stimulate its catalytic activity in vitro and in vivo [7–9]; and ( iii ) Dhh1 can detect reductions in ribosome transit at non-optimal codons and further inhibit translation elongation , eliciting acceleration of transcript decapping in vivo [6 , 10] . Based on their ability to inhibit translation initiation in vitro , it has been proposed that the decapping activators Dhh1 and Scd6 can direct their target mRNAs from a translationally active state to an mRNP state competent for mRNA storage or decapping [11] . Scd6 belongs to a highly conserved protein family with orthologs in humans ( hRAP55/Lsm14 ) , Xenopus laevis ( xRAP55 ) , Drosophila melanogaster ( Tral ) , Caenorhabditis elegans ( CAR-1 ) , Trypanosoma brucei ( TbSCD6 ) , and fission yeast ( Sum2 ) , among others [12] . All family members contain a conserved N-terminal LSm ( like-Sm ) domain , followed by central DFDF-FFD-TGF boxes , and variable numbers of C-terminal RGG/RGX motifs [12 , 13] . In metazoans and Plasmodium , Scd6 associates with Dhh1 homologs , other proteins , and mRNAs to form stable mRNPs that contain translationally silent transcripts subject to developmental regulation [12] . For example , in Drosophila , Scd6 and Dhh1 homologs , Tral and Me31B , belong to a repression complex containing CUP , an inhibitor of eIF4E-eIF4G association , shown to control the translation of certain mRNAs with key functions in embryogenesis [14 , 15] . This complex might function more broadly to repress translation of many mRNAs during early embryogenesis , possibly by coating the mRNA [16 , 17] . An Scd6 homolog in Xenopus , xRAPB , associates with the Dhh1 ortholog Xp54 in translationally inactive maternal transcripts in stored mRNPs; although xRAPB appears to oppose rather than promote Xp54 function in repressing translation in oocytes [18] . A longer Xenopus variant , xRAP55 represses translation in vitro , and decreases reporter protein levels when tethered to a reporter mRNA in cell extracts , dependent on its N-terminal LSm domain [19]; however , its role in oocytes is unclear [18] . In yeast and humans , Scd6 is required for the formation and accumulation of mRNA-containing cytoplasmic aggregates called Processing ( P ) bodies , and it also localizes to mRNA-containing stress granules under a variety of adverse conditions [19–22] . In addition to its ability to bind the decapping enzyme subunit Dcp2 , yeast Scd6 can also interact with the other decapping activators Pat1 and Edc3 [7 , 23–29] , as well as with various members of the Lsm complex and Dhh1 [23 , 24 , 29] . In vitro , yeast Scd6 represses translation initiation by directly interacting with the C-terminal region of eIF4G ( in the context of eIF4F ) via the Scd6 C-terminal RGG domain , preventing recruitment of the 43S PIC to activated mRNA and formation of the 48S PIC [7 , 22] . The RGG domain is also required for overexpressed Scd6 to inhibit cell growth and produce stress granules [22] . Yeast Scd6 can also interact with other translation components , including proteins of the small and large ribosomal subunits , PABP , eIF4B , eIF5 , and eEF1A [23 , 25 , 29] . Like yeast Scd6 , the Arabidopsis homolog , Dcp5 , was shown to repress translation in vitro [30] . The homolog in Trypanosoma brucei is present in cytoplasmic granules and appears to be a general repressor of translation , even though it does not exhibit an association with the Dhh1 homolog [13 , 16 , 17] that is otherwise conserved in yeast , worms , flies , and vertebrates [31] . Considerable evidence points to an important role for yeast Scd6 in cytoplasmic post-transcriptional control . It interacts with the decapping enzyme and its regulators , can inhibit 48S PIC assembly in vitro , and overexpression of its gene induces stress granule formation and inhibits cell growth in a manner requiring its C-terminal eIF4G-interaction domain . Nevertheless , it has not been demonstrated that Scd6 mediates the degradation or translational repression of any specific mRNA in yeast cells . In this study , we show that tethering Scd6 to two different reporter mRNAs in vivo evokes Dhh1-dependent repression of translation initiation , which is accompanied by reporter-specific Dcp2-mediated mRNA turnover . Both Scd6 functions are dependent on its LSm domain , and are enhanced by depletion of the Ccr4 deadenylase of the Ccr4-Not complex in vivo . By ribosome profiling we further provide evidence that Scd6 cooperates with Dhh1 in repressing the mRNA abundance and translation of particular native mRNAs in nutrient-replete medium , and present evidence that both functions require Dcp2 . Thus , Scd6 appears to be an important component of the gene expression control network in yeast cells , acting at the levels of both translation initiation and mRNA turnover .
Previously , it was shown that Scd6 interacts with mRNA decapping factors and can inhibit translation initiation in vitro [7 , 22]; however , it was unknown whether Scd6 can reduce the abundance and repress the translation of specific mRNAs in vivo . Because native mRNA substrates of Scd6 were unknown , we employed a tethered-function assay to examine the consequences of Scd6 binding to reporter mRNA—an approach used successfully to demonstrate regulation of mRNA translation and turnover by the decapping activator/translational repressor Dhh1 [6] . A fusion of Scd6 to bacteriophage MS2 coat protein ( CP ) , or MS2 CP alone , each tagged with three FLAG epitopes ( henceforth , Scd6-MS2-F or MS2-F ) , was expressed in wild-type ( WT ) cells from mRNA containing the 5’ and 3’ untranslated regions ( UTRs ) of SCD6 driven by the native SCD6 promoter from a low copy plasmid . A GFP reporter mRNA , harboring the 5’UTR and 3’UTR of PGK1 mRNA with tandem MS2 RNA recognition elements inserted in the 3’ UTR , and driven by a galactose-inducible GAL1 UAS and PGK1 hybrid promoter [6] , was expressed from a low copy plasmid in the same strains by culturing cells with galactose as carbon source . As a positive control , we expressed the Dhh1-MS2 fusion shown previously to evoke translational repression of the same GFP reporter [6] . Western blotting with anti-GFP antibodies and either Northern blotting or qRT-PCR analyses were conducted to measure steady-state expression of GFP protein and GFP reporter mRNA , respectively , and the ratio of these measurements yielded the translational efficiency ( TE ) of the reporter mRNA ( Fig 1A ) . Tethering Scd6-MS2-F reduced expression of GFP protein and reporter mRNA by comparable amounts , ~2 . 5- to 3-fold , compared to tethering MS2-F alone ( Fig 1B–1D , Scd6-MS2-F vs . MS2-F; P <0 . 0001 in C & D ) . These results are similar to those observed on tethering Dhh1-MS2 versus MS2 alone expressed from a high-copy plasmid ( Fig 1B–1D , Dhh1-MS2 vs . MS2 ( 2μM ) ) , in agreement with previous findings [6] . Expression of GFP protein and mRNA in cells expressing MS2-F was essentially indistinguishable from that measured in transformants of the same strain harboring empty vector ( S1A Fig , Fig 1C and 1D ) . Northern analysis confirmed the repression of GFP mRNA by both Scd6-MS2-F and Dhh1-MS2 in comparison to empty vector and MS2-alone controls ( Fig 1E ) . These results are consistent with the possibility that tethering Scd6 increases the rate of GFP mRNA turnover to lower its steady-state abundance , with attendant reduction in GFP protein expression . Owing to comparable repression of reporter protein and mRNA , tethered Scd6-MS2-F evokes little change in the TE of the GFP reporter ( Fig 1F ) . To determine whether tethering Scd6-MS2 to GFP mRNA accelerates its degradation , we measured the half-life of GFP mRNA following a shift from galactose to glucose medium that should repress the GAL promoter and halt new synthesis of reporter mRNA . On tethering Scd6-MS2-F , the half-life of GFP mRNA was significantly reduced from 2 . 8 ± 0 . 06 min ( tethering MS2-F alone ) to 1 . 9 ± 0 . 15 min ( P = 0 . 03; S1B Fig ) . A similar fold-reduction in half-life was observed on tethering Dhh1-MS2 vs MS2 alone ( 2 . 3 ± 0 . 09 min vs . 3 . 20 ± 0 . 02 min , respectively; P = 0 . 01; S1C Fig ) . We conclude that tethered Scd6-MS2-F reduces GFP mRNA abundance by accelerating its degradation . Because it was shown previously that Npl3 and Sbp1 can also repress translation initiation in vitro dependent on interaction of their RGG domains with eIF4G in a manner similar to Scd6 [22] , we examined MS2 fusions of these proteins , expressed from plasmid constructs containing the native promoters and 5’- and 3’-UTRs of NPL3 or SBP1 , respectively , along with the corresponding MS2 control proteins . In contrast to results obtained for Scd6-MS2-F , neither Npl3-MS2-F nor Sbp1-MS2-F had any significant effect on GFP expression from the same reporter analyzed above ( S2A , S2B , S2C and S2D Fig ) despite being expressed at levels exceeding that of Scd6-MS2–F ( S2E Fig ) . These results underscored the specificity of the effects of Scd6-MS2-F and led us to probe its mechanism of repression . As Scd6 can bind to Dcp2 , the catalytic component of the decapping enzyme [7] , we asked whether the repression of reporter mRNA abundance evoked by tethered Scd6-MS2-F is attenuated in a dcp2Δ strain . The isogenic WT DCP2 strain employed for this experiment exhibited reductions in GFP mRNA and GFP protein expression conferred by Scd6-MS2-F versus MS-F alone ( Fig 2A–2D , WT data ) equal to , or greater than those observed in the different WT strain employed above ( Fig 1B–1D ) . Tethering Scd6-MS2-F to the GFP reporter in dcp2Δ cells reduced the abundance of GFP mRNA ( Fig 2C , dcp2Δ , Scd6-MS2-F vs . MS2-F; P = 0 . 03 ) . Importantly , however , the reduction in reporter mRNA abundance conferred by Scd6-MS2-F was diminished in dcp2Δ versus WT cells , with the ratio of GFP mRNA in cells expressing Scd6-MS2-F versus MS2-F increasing from ~0 . 4 in WT to ~0 . 7 in dcp2Δ cells ( Fig 2D , white bars , WT vs . dcp2Δ; P = 0 . 009 ) . Expression of GFP protein in cells containing MS2-F alone was reduced somewhat in the dcp2Δ vs . WT strain; however , this reduction occurred independently of tethering as it was also observed in the corresponding strains containing empty vector ( S3A and S3B Fig ) . Tethering Scd6-MS2-F in dcp2Δ cells conferred a strong reduction in GFP protein expression compared to MS2-F alone ( Fig 2B , dcp2Δ , Scd6-MS2-F vs . MS2-F; P = 0 . 002 ) , that was only slightly smaller in magnitude compared to that seen in WT cells ( Fig 2D , black bars , WT vs . dcp2Δ ) . As a consequence of relatively greater repression of GFP protein versus GFP mRNA on tethering Scd6-MS2-F versus MS2-F alone in the dcp2Δ mutant ( Fig 2D , WT vs . dcp2Δ , black vs . white bars ) , the TE of GFP reporter mRNA is diminished in dcp2Δ versus WT cells by ~33% ( Fig 2E , WT vs . dcp2Δ; P = 0 . 005 ) . These findings suggest that: ( i ) Dcp2 is required for strong repression of reporter mRNA abundance , and ( ii ) translational repression of the GFP mRNA is unveiled in cells lacking Dcp2 , where mRNA turnover by tethered Scd6-MS2 is diminished . Similar conclusions were reached previously for tethered Dhh1-MS2 [6] . The concordance between effects on GFP reporter mRNA and protein expression conferred by tethered Scd6-SM-2 ( Fig 2A–2E ) and Dhh1-MS2 ( Sweet et al . , 2012 ) in WT and dcp2Δ cells raised the possibility that Dhh1 is required for translational repression conferred by tethered Scd6-MS2-F . To address this possibility , we examined the regulation of GFP reporter expression in dhh1Δ cells . In contrast to our findings for isogenic dcp2Δ cells , we observed greater repression of reporter mRNA abundance by tethered Scd6-MS2-F in dhh1Δ versus WT cells ( Fig 2C , Scd6-MS2-F vs . MS2-F , WT vs . dhh1Δ ) , with the Scd6-MS2-F/MS2-F repression ratio for GFP mRNA declining ~2 . 4-fold from ~0 . 4 in WT to ~0 . 17 in dhh1Δ cells ( Fig 2D , white bars , WT vs . dhh1Δ; P = 0 . 007 ) . Thus , unlike Dcp2 , Dhh1 is dispensable for repression of reporter mRNA abundance by tethered Scd6-MS2-F . One possible explanation for the enhanced repression of GFP mRNA in dhh1Δ cells ( vs . WT ) might be that Dhh1 impedes Dcp1/Dcp2-mediated decapping of the reporter in the presence of tethered Scd6-MS2-F . Repression of GFP protein by Scd6-MS2-F was also intact in the dhh1Δ strain ( Fig 2B , Scd6-MS2-F vs . MS2-F; P = 0 . 001 ) , with an Scd6-MS2-F/MS2-F repression ratio for GFP protein ~1 . 6-fold lower than that seen in WT cells ( Fig 2D , black bars , WT vs . dhh1Δ; P = 0 . 03 ) . However , owing to even greater Scd6-MS2-F-mediated repression of GFP mRNA compared to GFP protein , the TE of the GFP reporter was increased by ~1 . 4-fold in dhh1Δ vs . WT cells ( Fig 2E; P = 0 . 009 ) . This observation is consistent with the possibility that Dhh1 contributes to translational repression by tethered Scd6-MS2-F . We reasoned that if Dcp2 and Dhh1 are respectively required for repression of mRNA abundance and translation by tethered Scd6-MS2-F , then eliminating both proteins in a dcp2Δ dhh1Δ double mutant should abrogate repression of both GFP protein and GFP mRNA by Scd6-MS2-F . Indeed , nearly identical expression levels of GFP protein and mRNA were observed in the dcp2Δ dhh1Δ mutant whether expressing Scd6-MS2-F or MS2–F ( Fig 2B and 2C , dcp2Δ dhh1Δ , Scd6-MS2-F vs . MS2-F ) , yielding near-unity Scd6-MS2-F/MS2-F repression ratios in this strain for both reporter mRNA and protein expression ( Fig 2D , dcp2Δ dhh1Δ , black and white bars ) , which are markedly increased from the corresponding repression ratios of 0 . 36 and 0 . 40 in WT cells ( Fig 2D , dcp2Δ dhh1Δ vs . WT , white and black bars; P = 0 . 004 ( Protein ) , P = 0 . 0009 ( mRNA ) ) . Importantly , comparing the reporter TEs in the double mutant to the dcp2Δ single mutant reveals a loss of translational repression in dcp2Δ dhh1Δ cells ( Fig 2E , dcp2Δ dhh1Δ vs . dcp2Δ; P = 0 . 04 ) , supporting a requirement for Dhh1 in translational repression of GFP mRNA abundance by tethered Scd6-MS2-F . That the residual repression of mRNA abundance seen in the dcp2Δ single mutant is eliminated by dhh1Δ in the double mutant ( Fig 2D , white bars , dcp2Δ dhh1Δ vs . dcp2Δ; P = 0 . 02 ) might indicate that Dhh1 mediates a Dcp2-independent mechanism of mRNA degradation evoked by tethered Scd6-MS2-F in dcp2Δ cells , such as involvement of the exosome or a ribosome-associated endonuclease . In summary , analyses of dcp2Δ and dhh1Δ mutants indicate that efficient repression of reporter mRNA abundance is dependent on Dcp2 , whereas Dhh1 participates in translational repression , by tethered Scd6-MS2-F . Having observed that the occurrence of translational repression of GFP mRNA is unmasked in dcp2Δ cells , owing to partial stabilization of reporter mRNA , we examined the effect of tethered Scd6-MS2-F on the size distribution of GFP mRNA in polysomes , 80S monosomes , ribosomal subunits , and free mRNPs . Using sedimentation through sucrose density gradients and qRT-PCR analysis of GFP and actin ( ACT1 ) mRNA in the gradient fractions , we reproducibly observed a shift in the fractions with peak abundance of GFP mRNA from polysomes containing 3 to 6 ribosomes ( fractions 6–8 , 3- to 6-mers ) in cells expressing MS2-F alone to those containing 80S monosomes and 2- to 3-mers ( fractions 3–5 ) in cells expressing Scd6-MS2–F ( Fig 3A and 3B ( P-values for fractions 7 and 4 of 0 . 031 and 0 . 0005 ) ; S4A and S4B Fig ) . By contrast , the fractions containing the peak abundance of actin mRNA ( fractions 8–10 , 5- to 8-mers ) did not differ reproducibly between cells expressing Scd6-MS2-F versus MS2–F ( Fig 3A & 3C ) . Despite the shift in peak GFP mRNA abundance from 3- to 5-mers to 2-mers and monosomes , there was no reduction in the proportion of GFP mRNA found in the largest polysomes near the bottom of the gradient on tethering Scd6-MS2-F versus MS–F ( Fig 3A and 3B , fractions 10–15 ) . One way to explain these findings is to propose that tethering Scd6-MS2-F inhibits translation to a greater extent at the initiation versus elongation stage of protein synthesis for the fraction of GFP mRNA that shifts towards monosomes and free mRNP , while inhibiting elongation more than initiation on a minority fraction that is retained in heavier polysomes . Having implicated Dcp2 in the reduction of reporter mRNA abundance by tethered Scd6-MS2-F , and noting that mRNA degradation in yeast frequently proceeds via removal of the poly ( A ) tail followed by decapping [4] , we asked next whether repression of GFP mRNA expression by Scd6-MS2-F requires Ccr4 , the major cytoplasmic deadenylase in yeast [32] . Unexpectedly , the reduction in reporter mRNA abundance by Scd6-MS2-F was enhanced rather than diminished in cells lacking Ccr4 . We observed a modest ( ~30% ) reduction in GFP mRNA levels in the ccr4Δ mutant containing empty vector or MS2–F ( Fig 4B , vector & MS2-F , white vs . grey bars ) , but tethering Scd6-MS2-F conferred ~3-fold lower GFP mRNA abundance in ccr4Δ vs . WT cells ( Fig 4B , Scd6-MS2-F , white vs . grey bars ) . The resulting >5-fold reduction in mRNA expression by Scd6-MS2-F versus MS2-F seen in the ccr4Δ mutant ( Fig 4B , Scd6-MS2-F vs . MS2-F , white bars ) is >2-fold larger than that observed in WT cells ( Fig 4B , Scd6-MS2-F vs . MS2-F , grey bars; Fig 4C , ΔGFP mRNA , ccr4Δ vs . WT; P = 0 . 004 ) . These results suggest that Ccr4 is not only dispensable for the reduced abundance of GFP mRNA conferred by tethered Scd6-MS2-F , but actually appears to impede a more efficient repression pathway that can operate in its absence . Interestingly , the absence of Ccr4 also dramatically increased the repression of GFP protein expression by Scd6-MS2–F ( Fig 4A , Scd6-MS2-F vs . MS2-F , white vs . grey bars ) , reducing the Scd6-MS2-F:MS2-F repression ratio for GFP protein from 0 . 42 to ~0 . 07 ( Fig 4C , ΔGFP protein , ccr4Δ vs . WT; P = 0 . 0001 ) . Owing to greater repression of GFP protein versus GFP mRNA in ccr4Δ cells , tethering Scd6-MS2-F decreased the TE of GFP mRNA to ~40% of the corresponding TE of ~1 . 1 observed in WT cells ( Fig 4C , ΔTEGFP , ccr4Δ vs . WT; P<0 . 0001 ) . These findings imply that the presence of Ccr4 also interferes with a more efficient mechanism for translational repression by tethered Scd6-MS2-F that can proceed in ccr4Δ cells . Similar findings were observed on tethering Dhh1-MS2 versus MS2-F to GFP mRNA , as repression of protein expression was greatly enhanced in ccr4Δ versus WT cells ( Fig 4A and 4B , Dhh1-MS2-F vs . Scd6-MS2-F , grey vs . white bars ) , as previously observed with this same tethering system [6] . Recent findings suggest that the Caf1 subunit of the Ccr4-Not complex cooperates with Ccr4 in the deadenylation and degradation of a subset of yeast mRNAs with low codon optimality , functioning upstream of Dhh1-mediated decapping of such mRNAs [33] . Accordingly , we asked whether eliminating Caf1 from cells would diminish the repression of GFP mRNA abundance conferred by tethered Scd6-MS-F . At odds with this possibility , we observed a greater reduction in GFP mRNA abundance on tethering Scd6-MS2-F in caf1Δ compared to WT cells ( Fig 4D; P = 0 . 01 ) , similar to our findings with the ccr4Δ mutant ( Fig 4C ) . Thus , the Ccr4-Not complex is dispensable for , and even seems to impede , the degradation of mRNAs promoted by tethered Scd6-MS2-F , as observed previously for tethered Dhh1 [6 , 34] . To examine further whether tethered Scd6-MS2-F can repress translation of reporter mRNA independently of Ccr4 , we utilized an alternative reporter mRNA , in which the bacterial lacZ gene , encoding β-galactosidase , replaced the GFP coding sequences ( Fig 5A ) , and assays of β-galactosidase activity in cell extracts replaced Western analysis for quantifying reporter protein expression . As observed with the GFP reporter , tethering Scd6-MS2-F conferred an ~2- to 2 . 5-fold repression of β-galactosidase activity compared to that measured with MS2-F alone in the two different WT strains described above ( Fig 5B , Scd6-MS-F vs . MS2-F in WT ( BY4741 ) ( P<0 . 0001 ) and WT ( W303 ) ( P = 0 . 0005 ) ; Fig 5D , grey bars , WT strains ) . By contrast , β-galactosidase expression on tethering Npl3-MS2-F or Sbp1-MS2-F was indistinguishable from that observed for the corresponding MS2-only controls , or with empty vector in WT cells ( S5C Fig ) . Thus , tethering Scd6-MS2-F , but not the corresponding Npl3 or Sbp1 fusions , confers similar repression of protein expressed from GFP and lacZ reporters . Repression of the lacZ reporter by Scd6-MS2-F was intact in an scd6Δ strain ( S5A Fig; P<0 . 0001 ) , ruling out a contribution of native Scd6 to the function of tethered Scd6-MS2-F . Expression of β-galactosidase activity from heterologous GCN4-lacZ or GAL1-lacZ reporters lacking MS2 binding sites was indistinguishable in cells expressing Scd6-MS-F or MS2–F ( S5B Fig ) , indicating that repression by Scd6-MS-F requires its tethering to lacZ mRNA . Interestingly , lacZ reporter mRNA abundance was not significantly altered , or reduced by only ~20% , by tethered Scd6-MS2-F in the two different WT strains ( Fig 5C , Scd6-MS-F vs . MS2-F in WT ( BY4741 ) and WT ( W303 ) ( P = 0 . 02 ) ; Fig 5D , ΔlacZ mRNA , WT strains ) . As a consequence of greater repression of protein versus mRNA expression in WT cells ( Fig 5D , WT strains , grey vs . white bars ) , tethering Scd6-MS2-F reduced the TE of lacZ reporter mRNA by ~30–50% compared to MS2-F alone in the WT strains ( Fig 5E , WT strains ) . Thus , unlike our findings for the GFP reporter ( Fig 1F , ΔTEGFP ) , the ability of tethered Scd6-MS2-F to repress translation of the lacZ reporter was observable in WT cells containing Dcp2 . Tethering Scd6-MS2-F in ccr4Δ cells conferred a dramatic ~9-fold reduction in β-galactosidase expression , and also a ~2 . 7-fold decrease in lacZ mRNA expression , in comparison to MS2-F alone ( Fig 5B and 5C , Scd6-MS-F vs . MS2-F , ccr4Δ; P = 0 . 001 ( panel B ) , P<0 . 0001 ( panel C ) . The reductions in β-galactosidase expression and lacZ mRNA on tethering Scd6-MS-F were both greater in ccr4Δ versus WT ( BY4741 ) cells ( Fig 5D , grey and white bars , ccr4Δ vs . WT ( BY4741 ) ; P = <0 . 0001 ( grey ) , P = 0 . 0005 ( white ) . However , the reduction in β-galactosidase was relatively larger , yielding an ~3-fold reduction in TE attributable to tethered Scd6-MS2-F , which exceeds the ~2-fold reduction in TE found in WT cells ( Fig 5E , ccr4Δ; P = 0 . 002 ) . Thus , eliminating Ccr4 enhances the repression of both mRNA abundance and translational efficiency of the lacZ reporter by tethered Scd6-MS2-F , as observed above for the GFP reporter ( Fig 4C ) . The finding that tethered Scd6-MS2-F yields a larger reduction in GFP versus lacZ reporter mRNA abundance in WT cells might be explained by proposing that another step in mRNA turnover that is not accelerated by tethered Scd6-MS-F ( eg . exonucleolytic degradation by Xrn1 or deadenylation ) is more rate-limiting than decapping for the lacZ reporter . Compared to the levels observed in the isogenic WT strain , the levels of the lacZ reporter mRNA and expression of β-galactosidase were substantially reduced in dhh1Δ cells , even in the presence of empty vector or MS2–F ( S5E and S5F Fig , vector and MS2-F , grey vs . white bars ) . Expression of an unrelated GAL1-lacZ fusion was also diminished by dhh1Δ in cells lacking any MS2 proteins , albeit to a smaller degree ( S5D Fig ) . While these effects of dhh1Δ complicated our analysis , we nevertheless obtained results consistent with the earlier conclusion that Dhh1 is required for translational repression by tethered Scd6-MS2-F . Despite the general reductions in lacZ mRNA levels in dhh1Δ cells , Scd6-MS2-F conferred ~2 . 5-fold lower mRNA levels versus MS2-F alone in this mutant ( Fig 5C , Scd6-MS2-F vs . MS2-F , dhh1Δ; P = 0 . 0001 ) . Thus , eliminating DHH1 uncovers a reduction in lacZ mRNA abundance on tethering Scd6-MS2-F not observed in the isogenic WT strain ( Fig 5D , white bars , dhh1Δ vs . WT; P = 0 . 0002 ) . Expression of β-galactosidase in dhh1Δ cells also showed an ~2-fold reduction on tethering Scd6-MS2-F versus MS2–F ( Fig 5B , Scd6-MS2-F vs . MS2-F , dhh1Δ; P<0 . 0001 ) , as in the WT strain ( Fig 5D , white grey bars , dhh1Δ vs WT ) . Because tethered Scd6-MS2-F reduced both reporter mRNA and reporter protein by ~60% compared to MS2-F alone ( Fig 5D , dhh1Δ ) , it did not confer any reduction in TE of lacZ mRNA , in contrast to the ~50% reduction in TE observed in WT cells ( Fig 5E , dhh1Δ vs . WT; P = 0 . 001 ) , which is consistent with Dhh1 being required for translational repression of the lacZ reporter . Deletion of DCP2 led to an ~2 . 5-fold increase in lacZ mRNA levels in the presence of MS2-F alone ( Fig 5C , MS2-F , dcp2Δ vs . WT ( W303 ) ) , which also occurred in the presence of vector alone and , hence , is not a consequence of tethering ( S5G Fig ) . Tethering Scd6-MS2-F evoked an ~27% additional increase in lacZ reporter mRNA versus tethering MS2-F in dcp2Δ cells ( Fig 5C , Scd6-MS2-F vs . MS2-F , dcp2Δ; P = 0 . 03 ) , eliminating the small reduction in lacZ mRNA abundance on tethering Scd6-MS2-F in the isogenic WT strain noted above ( Fig 5C , Scd6-MS2-F vs . MS2-F , WT; Fig 5D , white bars , dcp2Δ vs . WT; P = 0 . 001 ) . The fact that the ~1 . 3-fold increase in lacZ mRNA conferred by tethered Scd6-MS2-F versus MS2-F in dcp2Δ cells is smaller than the corresponding ~2-fold increase in GFP mRNA abundance conferred by tethered Scd6-MS2-F in dcp2Δ versus WT cells shown above ( Fig 2C , black vs . dark grey bars ) might reflect that tethering Scd6-MS2-F confers a much smaller reduction in lacZ mRNA ( ~1 . 1-fold ) versus GFP mRNA abundance ( ~2 . 6-fold ) in WT cells ( Figs 5D vs . 2D , WT cells , white bars ) . Coupling the small increase in lacZ mRNA abundance with a slight reduction in β-galactosidase expressed on tethering Scd6-MS2-F versus MS2-F in dcp2Δ cells ( Fig 5B and 5C , Scd6-MS2-F vs . MS2-F , dcp2Δ ) , results in an ~30% reduction in TE compared to tethering MS2-F alone , which is indistinguishable from that observed in the isogenic WT strain ( Fig 5E , dcp2Δ vs . WT ( W303 ) ) . These findings are consistent with our conclusion above that Dcp2 is dispensable for translational repression by tethered Scd6-MS2-F . In summary , tethering Scd6-MS2-F to the lacZ reporter confers a decrease in translational efficiency that appears to be dependent on Dhh1 , independent of Dcp2 , and dampened by Ccr4; and Ccr4 also diminishes the repression of lacZ mRNA abundance by tethered Scd6-MS2-F . All of these observations are in agreement with our findings for the GFP reporter . Unlike our results for the GFP reporter , where deleting DCP2 reduced the TE on tethering Scd6-MS2–F ( Fig 2E , cols . 1–2 ) , dcp2Δ had little effect on TE of the lacZ reporter because it did not substantially reduce the apparent degradation of lacZ mRNA by tethered Scd6-MS2-F , conferring only a small increase in lacZ mRNA abundance ( Fig 5D , WT vs . dcp2Δ , white bars . ) Scd6 interacts with eIF4G via the C-terminal region of Scd6 containing the RGG domain [22] . Interaction partners of the N-terminal LSm domain of Scd6 are unknown; however , the LSm domains in Scd6 homologs from different species mediate binding to Dcp2 in S . pombe [35] , both Dcp1 and the translational repressor CUP in D . melanogaster [15] , and decapping activators and translational repressors 4E-T and EDC4 in humans [36] . We examined the importance of the LSm and RGG domains of Scd6 for reporter mRNA repression by truncating the Scd6-MS2-F fusion at the N- or C-terminal ends to remove these domains individually ( Fig 6A ) . Eliminating the LSm domain completely abrogated repression of both protein and mRNA expressed from the GFP reporter by tethered Scd6-MS2–F ( Fig 6D and 6E , white vs . grey bars; P = 0 . 0001 ( D ) , P<0 . 0001 ( E ) for ΔLSm-Scd6-MS2-F vs . Scd6-MS2-F ) , without detectably altering expression of the Scd6-MS2-F fusion protein ( Fig 6B , upper blot , lanes 5–12 ) . The fact that no reduction in GFP mRNA occurs on tethering the ΔLSm-Scd6-MS2-F variant implies that the LSm domain is required for accelerated mRNA turnover conferred by WT tethered Scd6-MS2-F . Because there is no reduction in GFP protein expression on tethering the ΔLSm-Scd6-MS2-F , despite high levels of the GFP reporter mRNA , we can also infer that translational repression of GFP mRNA is eliminated by removing the LSm domain . Repression of β-galactosidase expression from the lacZ reporter was also abolished by removing the LSm domain from Scd6-MS2–F ( Fig 6G , white vs . grey bars; P<0 . 0001 for ΔLSm-Scd6-MS2-F vs . Scd6-MS2-F ) . Because there is little or no reduction in lacZ mRNA abundance on tethering WT Scd6-MS2-F , it seems likely that translational repression is abrogated by removing the LSm domain for this reporter mRNA as well . ( However , we cannot discard the unlikely possibility that tethering ΔLSm-Scd6-MS2-F would substantially increase the abundance of lacZ mRNA and thereby mask efficient translational repression by the ΔLSm variant . ) In contrast to our findings on removing the LSm domain , eliminating the RGG domain from Scd6-MS2-F had no apparent effect on repression of the GFP reporter by Scd6-MS2–F ( Fig 6C , GFP blot; Fig 6F , black vs . grey; P = 0 . 002 ) , although an apparent increase in expression of the ΔRGG-Scd6-MS2-F versus WT Scd6-MS2-F fusion ( Fig 6C , upper blot , lanes 5–12 ) might have obscured a reduced efficiency of reporter repression for the ΔRGG variant . These findings indicate that the LSm domain , and most likely interactions it mediates with effector proteins , is required for both enhanced degradation of GFP reporter mRNA and translational repression of both reporters by tethered Scd6-MS2-F . By contrast , interaction of Scd6-MS2-F with eIF4G via the Scd6 RGG domain might be dispensable for translational repression when Scd6 is tethered tightly to the mRNA; although we cannot eliminate the possibility that translational repression is impaired by the ΔRGG truncation and that the efficient repression of GFP reporter protein expression conferred by ΔRGG-Scd6-MS2-F occurs exclusively from accelerated mRNA turnover . To determine whether Scd6 and Dhh1 participate in regulating the abundance or translation of native yeast mRNAs , we conducted ribosome footprint profiling and RNA-Seq analyses on the WT , dcp2Δ , scd6Δ and dhh1Δ strains in the W303 genetic background , cultured in rich ( YPD ) medium . We also analyzed isogenic dcp2Δscd6Δ and dcp2Δdhh1Δ double mutants , anticipating that changes in translational efficiency might be more evident in the absence of mRNA decapping by Dcp2 . Independent RNA-Seq analysis was also conducted in parallel on two isogenic scd6Δ strains , an additional isogenic dhh1Δ strain and isogenic mutants lacking the Dcp2-decapping activators Pat1 and Lsm1 [4] to determine whether Pat1 and Lsm1 contribute to Scd6-mediated repression of mRNA abundance . The results of biological replicates were highly correlated for both ribosome-protected fragments ( RPFs ) and mRNA sequences for all strains analyzed ( S6A–S6L Fig ) . RNA-Seq analysis of WT and scd6Δ strains identified 83 mRNAs whose abundance was significantly up-regulated in the mutant by ≥1 . 4-fold at an FDR of <0 . 01 , with a median fold-change ( FC ) compared to WT of ~1 . 8 , which is significantly higher than the median FC for all mRNAs ( which is 1 . 0 ( log2 ( ΔmRNA ) = 0 ) ) owing to normalization for equal RNA read numbers for all genes in each strain ) ( Fig 7A , scd6Δ ) . Interestingly , this group of mRNAs also displayed significantly increased abundance in the isogenic dhh1Δ , pat1Δ , and lsm1Δ strains , with median FCs of ~3 . 3 , ~1 . 9 , and ~1 . 8 , respectively ( Fig 7A , cols . 2–4 ) . Consistent with the latter , most of the 83 mRNAs up-regulated in scd6Δ cells are a subset of the larger group of 733 mRNAs elevated to the same degree in dhh1Δ cells ( Fig 7B ) . Importantly however , derepression of these 83 mRNAs was not observed on deleting SCD6 or DHH1 in the strain lacking DCP2 , ie . when comparing the dcp2Δscd6Δ and dcp2Δdhh1Δ double mutants to the dcp2Δ single mutant ( Fig 7A , rows 5–6 ) . These findings are consistent with the notions that: ( i ) Scd6 accelerates degradation of a subset of native mRNAs , ( ii ) that Dhh1 , Pat1 , and Lsm1 all participate in this down-regulation of mRNA abundance , and ( iii ) that the decapping enzyme subunit Dcp2 is required for both Scd6- and Dhh1-enhanced degradation of the set of Scd6-targeted mRNAs . We obtained complementary results for a group of 346 mRNAs whose abundance was significantly increased ( at FDR<0 . 01 ) in an isogenic dhh1Δ mutant , exhibiting an ~3-fold increase in median mRNA abundance in dhh1Δ vs . WT cells ( S7A Fig , dhh1Δ ( z ) ) , and also showing ~1 . 9-fold and ~1 . 8-fold increases in two published RNA-seq datasets for a dhh1Δ mutant in the BY4741 background [10 , 37] ( S7A Fig , dhh1Δ ( r ) and dhh1Δ ( j ) ) . This group of Dhh1 down-regulated mRNAs also displays a slight , but statistically significant , ~15% derepression in the scd6Δ mutant ( S7A Fig , scd6Δ ) . Supporting these findings , hierarchical clustering analysis of expression changes for all mRNAs revealed that a large proportion of genes exhibit altered mRNA levels in the same direction in response to scd6Δ or dhh1Δ ( S7B Fig ) , which is particularly evident for the mRNAs showing the largest fold-changes in dhh1Δ vs . WT cells ( S7C Fig ) , with the magnitudes of these changes being generally greater in dhh1Δ vs . scd6Δ cells ( S7B and S7C Fig ) . These results suggest that Scd6 contributes appreciably to Dhh1-enhanced degradation of a large fraction of the mRNAs whose abundance is repressed by Dhh1 . As observed for the Scd6 down-regulated mRNAs , the derepression of mRNA levels for the group of 346 Dhh1-repressed mRNAs conferred by either scd6Δ or dhh1Δ was eliminated when these mutations were made in cells lacking DCP2 ( S7A Fig , cf . cols . 1–2 and 5–6 ) , indicating that Dcp2 is required for robust Dhh1-mediated mRNA turnover . The two groups of 83 and 346 mRNAs whose abundance is derepressed in scd6Δ ( Fig 7A ) or dhh1Δ cells ( S7A Fig ) , respectively , were interrogated next for changes in translation efficiency ( TE ) in different mutants by combining the results of ribosome profiling and RNA-Seq experiments conducted on the same cell cultures . The TE of each mRNA was calculated as the sum of RPFs divided by the sum of RNA reads across the CDS , and the change in TE ( ΔTE ) was calculated as the ratio of TE in mutant versus WT cells . ( Because RPFs and mRNA reads for each gene are normalized to RPF or mRNA reads for all genes in each strain , the ΔTE for each gene is determined relative to the median ΔTE for all genes , which is ~1 . 0 ) . Both groups of mRNAs showing derepression of mRNA abundance in scd6Δ or dhh1Δ cells also exhibited modest increases in median TE in our dhh1Δ mutant of ~12–15% ( S7D and S7E Fig , col . 5 in each panel ) . Comparable or somewhat greater increases in TE were identified in the two published ribosome profiling/RNA-seq datasets for a dhh1Δ mutant in the BY4741 background [10 , 37] ( S7D and S7E Fig , cols . 3–4 in each panel ) ; whereas the scd6Δ mutation increased the median TE for these groups of mRNAs by only ~5% ( S7D and S7E Fig , col . 1 ) . Despite the modest TE increases for these groups of mRNAs conferred by dhh1Δ , it is noteworthy that these changes were not observed on comparing the dcp2Δdhh1Δ double mutant to the dcp2Δ single mutant ( S7D and S7E Fig , cols . 5–6 in each panel ) , providing genetic evidence that the TE changes are genuine , and indicating that translational repression by Dhh1 is dependent on Dcp2 . Evidence for Dcp2-dependent translational repression of native mRNAs by Scd6 was provided by examining a group of 78 mRNAs exhibiting significantly increased ribosome occupancies in the scd6Δ mutant , and for a second group of 53 mRNAs showing the largest TE increases conferred by scd6Δ . In response to scd6Δ , both groups of mRNAs exhibit increased median TEs of ~1 . 2- and ~1 . 5-fold , respectively , in otherwise WT cells , but not in the presence of dcp2Δ ( Fig 7C & 7E , cf . cols . 1–2 in each panel ) . Comparable , or somewhat greater , increases in median TE were observed for both groups of mRNAs in all three dhh1Δ datasets ( Fig 7C & 7E , cols . 3–5 ) , indicating that Dhh1 contributes to translational repression of a substantial proportion of the mRNAs thus repressed by Scd6 . This last inference is further supported by the significant overlaps between mRNAs exhibiting increased ribosome occupancies or TEs in response to scd6Δ and the larger groups of mRNAs showing comparable increases in ribosome occupancy or TE in response to dhh1Δ ( Fig 7D & 7F ) . Moreover , clustering analysis for all mRNAs showed that the majority of mRNAs displaying increased TEs in response to scd6Δ also exhibit TE increases of generally greater degree in response to dhh1Δ; although it is also evident that many mRNAs translationally repressed by Dhh1 are not repressed by Scd6 ( Fig 8A and 8B ) . Once again , comparing the dcp2Δdhh1Δ double mutant to the dcp2Δ single mutant revealed that dcp2Δ largely suppresses the TE increases conferred by dhh1Δ in DCP2 cells ( Fig 7C & 7E , cf . cols . 5–6 in each panel ) , supporting a widespread requirement for Dcp2 in translational repression by Dhh1 . The mRNAs encoded by YOR173W , YHR033W , YGR088W , YFR017C , and YNR034W-A exhibit TE increases conferred by scd6Δ , and they respond similarly to dhh1Δ , while exhibiting larger increases in mRNA and TE levels in dhh1Δ versus scd6Δ cells ( Fig 8C and S8A–S8D Fig ) . Gene-ontology ( GO ) analysis revealed that mRNAs whose transcript abundance was derepressed in dhh1Δ or scd6Δ cells are enriched in related functional categories of genes involved in metabolism of energy reserves and other aspects of carbohydrate metabolism; and the Dhh1-repressed mRNAs are also enriched in stress-response genes ( S9A and S9B Fig ) . These results suggest that Scd6 and Dhh1 cooperate in repressing both mRNA abundance and translational efficiency for a discrete subset of native mRNAs in nutrient-replete medium . To provide evidence supporting a direct role for Dhh1 in translational repression of native mRNAs by Scd6 , we interrogated published results from deep-sequencing of mRNAs specifically immunoprecipitated with epitope-tagged Dhh1 ( RIP-Seq ) from WT cells grown on rich medium [38] . The 3686 mRNAs for which both RIP-Seq and ribosome profiling data exist were sorted into 5 equal percentiles based on Dhh1 enrichment values in RIP-seq and compared for translation changes in response to dhh1Δ or scd6Δ . This analysis revealed a direct correlation between Dhh1 enrichment values and changes in both ribosome occupancies and TEs in response to dhh1Δ ( S10A and S10B Fig ) , and a similar correlation exists between Dhh1 enrichment and changes in mRNA levels in dhh1Δ vs . WT cells ( S10C Fig ) , as noted previously [38] . These correlations support previous conclusions that Dhh1 binding to mRNAs is associated with accelerated mRNA decay and translational repression . Interestingly , Dhh1 enrichment values are likewise correlated with changes in ribosome occupancies and TEs in response to scd6Δ ( Fig 8D and 8E ) , supporting the notion that Scd6 translational repression of many native mRNAs involves recruitment of Dhh1 . Finally , we considered the possible contribution of codon optimality in dictating susceptibility of mRNAs to Scd6 . Previously , low codon optimality was associated with Dhh1-mediated mRNA decay in part by demonstrating an inverse correlation between the sTAI value , a measure of overall codon optimality of the mRNA [39] , and the change in mRNA abundance in dhh1Δ versus WT cells [10] . This correlation is also evident in our dhh1Δ dataset and that of Jungfleisch et al . [37] ( S11B and S11C Fig ) , although less pronounced than observed in the data from Radhakrishnan et al . [10] ( S11A Fig ) . A similar , modest trend was also evident for mRNA changes observed here in scd6Δ cells ( S12A Fig ) , suggesting that mRNAs exhibiting Scd6-dependent mRNA degradation have a tendency to exhibit poor codon optimality . Interestingly , however , the group of 82 mRNAs whose abundance is most strongly derepressed in scd6Δ cells ( characterized in Fig 7A ) exhibit sTAI values that are somewhat higher , not lower , than the genome average value ( S12B Fig ) , indicating that poor codon optimality is not the key determinant of Scd6-dependent mRNA turnover for the transcripts that it most strongly represses .
In this report we have shown that tethering an Scd6-MS2-F fusion protein to two different reporter mRNAs harboring MS2 binding sites represses reporter protein expression and , in the case of the GFP reporter , also reduces reporter mRNA abundance . Together with Npl3 and Sbp1 , Scd6 is one of three yeast proteins containing RGG domains capable of binding to the C-terminal domain of eIF4G and repressing translation initiation in cell extracts [22] . However , we observed no effects on reporter expression on tethering Npl3 or Sbp1 . The fact that tethered Scd6-MS2 reduced β-galactosidase expression without reducing lacZ mRNA abundance implied a reduction in translational efficiency of the lacZ mRNA . While this inference was not possible for the GFP reporter in WT cells , owing to comparable reductions in protein and mRNA expression , it was clearly indicated by the much greater repression of GFP protein versus GFP mRNA conferred by tethered Scd6-MS2 in the ccr4Δ mutant . Moreover , translational repression of the GFP reporter was revealed in the dcp2Δ mutant , as the reduction in GFP mRNA abudance was diminished while repression of GFP protein was maintained . Thus , we propose that tethering Scd6-MS2 evokes translational repression of both GFP and lacZ reporter mRNAs , and also degradation of the GFP reporter mRNA , with the latter dependent on decapping by Dcp2 . Additional evidence for translational repression was provided by our finding that tethering Scd6-MS2 in dcp2Δ cells shifted a proportion of the GFP mRNA from large to smaller polysomes and monosomes , suggesting a reduced rate of translation initiation . We implicated Dhh1 in translational repression of both GFP and lacZ mRNAs , but found it to be dispensable for the degradation of GFP mRNA , evoked by tethered Scd6-MS2-F . Thus , the TE values for both GFP and lacZ mRNAs were higher in dhh1Δ versus WT cells ( Figs 2E & 5E ) ; and also were higher in the dhh1Δ dcp2Δ double mutant compared to the dcp2Δ single mutant for the GFP reporter ( Fig 2E ) . Importantly , repression of both GFP protein and GFP mRNA abundance by tethered Scd6-MS2 is absent in the dhh1Δ dcp2Δ double mutant , whereas GFP protein repression is intact in dcp2Δ cells , and repression of GFP mRNA abundance occurs in dhh1Δ cells ( Fig 2D ) . These comparisons indicate that Dcp2 is required for efficient mRNA degradation while Dhh1 is required for full translational repression of GFP reporter mRNA . The role of Dhh1 in translational repression is further supported by our finding that tethering Dhh1 as an MS2 fusion represses expression of GFP protein more than GFP mRNA abundance in ccr4Δ cells , as observed previously [6] ) , similar to the effects of tethered Scd6-MS2–F ( Fig 4A and 4B ) . These results are consistent with previous demonstrations of direct interactions between Scd6 and Dhh1 [23 , 24] . Our finding that repression of GFP mRNA abundance by Scd6-MS2 remains intact in the dhh1Δ strain implies that Dhh1 is not required for recruitment or activation of Dcp1/Dcp2 on this reporter mRNA , which could involve instead the known direct interaction of Scd6 with Dcp2 [7 , 23 , 24 , 27 , 29] . Although we could readily observe that tethered Scd6-MS2-F reduces the TE of the GFP reporter mRNA in ccr4Δ cells , it was not possible to infer translational repression in WT cells because tethered Scd6-MS2-F repressed GFP reporter and protein expression almost equally . However , translational repression by tethered Scd6-MS2-F was uncovered in the dcp2Δ strain in which the accelerated degradation of GFP reporter mRNA was diminished . The ability to observe translational repression after uncoupling it from mRNA turnover in a dcp2Δ mutant has been reported previously for Dhh1 using the same tethering assay and GFP reporter employed here [6] , and also in similar experiments involving the tethering of Dhh1 to reporter mRNA in mutant strains where mRNA turnover was impaired by elimination of Dcp1 or Xrn1 [34] . Together , these findings indicate that both Scd6 and Dhh1 can repress translation independently of their functions in activating mRNA decay . In addition to the genetic uncoupling of mRNA decay from translational repression accomplished in yeast , these processes have been kinetically resolved in miRNA-mediated repression in Drosophila cells [40] and zebrafish [41] by showing that translational repression precedes mRNA turnover . Previous findings showed that tethered Dhh1-MS2 interferes with the elongation stage of translation and is associated with the presence of slowly decoded suboptimal codons in the reporter mRNA [6] . Current evidence suggests that Dhh1 can be recruited to mRNAs by slowly elongating ribosomes and triggers decapping and subsequent mRNA degradation; and that Dhh1 can also impede the progression of 80S ribosomes at suboptimal codons , at least when tethered to mRNA or overexpressed in cells [10] . These findings are ostensibly at odds with our conclusion that Dhh1 participates in translational repression by tethered Scd6-MS2-F and the results of our polysome analysis indicating that tethered Scd6-MS2 does not shift the GFP reporter mRNA into larger polysomes ( Fig 3 ) , which would be expected for slower elongation . Rather , tethered Scd6-MS2 appears to shift the mRNA towards smaller polysomes and possibly free mRNP ( Fig 3 ) . However , Dhh1 can bind directly to both 40S and 60S subunits [6 , 10] , and has been implicated in the inhibition of bulk translation initiation during carbon starvation [42] or when overexpressed in nutrient-replete cells [9] . Moreover , Dhh1 can inhibit 48S PIC assembly when added to cell extracts [9] . It has been suggested that Dhh1 can inhibit either initiation by interacting with the PIC , or elongation by binding to translating 80S ribosomes , and the relative importance of these mechanisms could vary with the mRNA , depending , for example , on the number and position of suboptimal codons [6 , 10] . Although tethered Dhh1-MS2 was found to inhibit elongation on the GFP reporter mRNA [6] , for which tethered Scd6-MS2 appears to have a relatively greater effect on initiation , perhaps the amount of Dhh1 that would be recruited to the reporter by Scd6-MS2 is lower than achieved by tethering Dhh1-MS2 itself , and may be sufficient to inhibit initiation but not elongation . As shown for Dhh1 , Scd6 can also inhibit 48S PIC formation in cell extracts [7 , 22] and tethered Scd6-MS2-F might work in conjunction with Dhh1 to produce a rate-limiting initiation defect on the GFP reporter . We found that the LSm domain of Scd6 is indispensable for the ability of tethered Scd6-MS-F to repress GFP reporter mRNA and protein expression ( Fig 6 ) , implying its requirement for both translational repression by Dhh1 and decapping by Dcp2 . Whereas the LSm domains of Scd6 homologs in S . pombe and D . melanogaster have been shown to interact with Dcp1 or Dcp2 , the interactions with Dhh1 homologs involve the DFDF and TFG domains in the C-terminal regions of Scd6 homologs in D . melanogaster [15] and humans [36] . If the LSm domain in S . cerevisiae Scd6 likewise interacts with Dcp1 or Dcp2 , this could explain the requirement for this domain in stimulating mRNA degradation by tethered Scd6-MS2-FL; however , the additional requirement of the LSm domain for translational repression presumably does not involve direct recruitment of Dhh1 . Considering that the LSm domains in Scd6 homologs also bind the translational repressor proteins CUP in Drosophila [15] and 4E-T in humans [36] , we suggest that this domain in S . cerevisiae Scd6 likewise recruits an additional repressor protein that , together with Dhh1 , mediates translational inhibition by tethered Scd6-MS2-FL . In contrast to our findings on the LSm domain of Scd6 , we found that the RGG domain at the C-terminus of Scd6 was not needed for repression of the GFP reporter by tethered Scd6-MS-F . As interaction of the Scd6 RGG domain with eIF4G was found previously to be required for the inhibition of translation initiation by Scd6 in cell extracts [22] , it is possible that this interaction interferes with the intrinsic function of the eIF4G C-terminal region in 48S PIC assembly . Alternatively , the Scd6-RGG/eIF4G interaction could serve primarily to recruit Scd6 to eIF4F-mRNP complexes for inhibition of 43S PIC recruitment , via Scd6 interactions with other components of the eIF4F-mRNP or 43S PIC , or by recruiting repressor proteins like Dhh1 or Pat1 to do so . Although our findings are more consistent with the latter possibility , an inhibitory interaction of the Scd6 RGG domain with the eIF4G C-terminus might still be crucial for translational repression on native mRNAs to which Scd6 is not tightly tethered . Our results using the tethering assay demonstrate that Scd6 can repress both mRNA abundance and translational efficiency of specific reporter mRNAs when tethered to these mRNA targets in vivo . Using RNA-Seq and ribosome profiling we went on to provide evidence that Scd6 is involved in repressing the abundance and/or translation of a discrete set of native yeast mRNAs in cells cultured in rich medium . The abundance of a group of 83 mRNAs was significantly up-regulated in scd6Δ cells , and the functions of the encoded proteins are enriched in the processes of metabolism of energy reserves and other aspects of carbon metabolism . Similarly , the abundance of a group of 346 mRNAs was derepressed in dhh1Δ cells , which are enriched for the same functional categories , as well as in stress response functions . These findings are consistent with recent results indicating that Dhh1-occupied mRNAs are enriched for transcripts whose levels are derepressed in cells depleted of glucose or a preferred nitrogen source [38] . Importantly , the group of 83 mRNAs whose levels are elevated in scd6Δ cells also tend to be elevated in mutants lacking Dhh1 , Pat1 , or Lsm1 ( Fig 7A and 7B ) , suggesting cooperation among these decapping activators in degradation of many native Scd6 target mRNAs . The groups of mRNAs whose abundance is derepressed in scd6Δ or dhh1Δ cells also exhibit a modest up-regulation in median TE values in dhh1Δ cells ( S7D and S7E Fig ) , consistent with concerted mRNA destabilization and translational repression by Dhh1 on a subset of these mRNAs . It is possible that the observable extent of translational repression of these mRNAs is dampened by their accelerated degradation , in the manner we observed for the GFP reporter mRNA on tethering Scd6-MS2-F . For two additional groups of mRNAs exhibiting the largest increases in ribosome occupancy or TE in scd6Δ cells , we again observed a contribution of Dhh1 to translational repression ( Fig 7C–7F ) , similar or even greater in magnitude to that of Scd6 for these groups of mRNAs ( Fig 7C & 7E ) . Broad cooperation between Scd6 and Dhh1 in translational control was also evident in genome-wide comparisons of TE changes in scd6Δ vs . dhh1Δ cells ( Fig 8A and 8B ) . Moreover , we found that Dhh1 occupancy is correlated with increased translation and increased TE in scd6Δ cells ( Fig 8D and 8E ) as well as in dhh1Δ cells ( Fig 9A and 9B ) . These findings support our conclusion reached from tethering assays that translational repression by Scd6 involves Dhh1 . Interestingly , our analyses of double mutants lacking Dcp2 in addition to Scd6 or Dhh1 indicated that translational repression , as well as mRNA degradation , mediated by Scd6 or Dhh1 is highly dependent on Dcp2 for most native mRNAs regulated by these proteins . Dcp2-dependence was expected for repression of mRNA levels , as decapping is an established prelude to mRNA degradation by the 5’-3’ exonuclease Xrn1 in yeast [4] . We did not anticipate a requirement for Dcp2 in translational repression , however , as translational repression of the GFP reporter by both tethered Scd6-MS-F ( Fig 2 ) and Dhh1-MS [6] was uncovered in dcp2Δ cells by the reduced GFP mRNA turnover , rather than being diminished . One way to account for this discrepancy is to propose that , on native mRNAs targeted by Scd6 or Dhh1 , Dcp2 is required for stable assembly of a translation repression complex capable of impeding 43S PIC association ( Fig 9 , ( iii ) ) , in addition to decapping mRNA to enhance degradation ( Fig 9 , ( iv ) ) , and this contribution of Dcp2 to translational inhibition is bypassed by artificially increasing the occupancies of Scd6 or Dhh1 on the mRNAs via tethering . It is also possible that a broad effect of dcp2Δ in increasing the abundance of many capped mRNAs , possibly with shortened poly ( A ) tails , indirectly diminishes translational repression by Dhh1 or Scd6 binding to target mRNAs . Although the underlying mechanism for the role of Dcp2 in translational repression of native mRNAs remains to be determined , the fact that dcp2Δ completely suppressed the increased mRNA levels and TE values conferred by scd6Δ or dhh1Δ ( Fig 7A , 7C and 7E ) provides genetic evidence that , while modest in magnitude for scd6Δ , these changes are physiologically relevant for the affected mRNAs . An unexpected finding from the tethering assays is that Scd6-MS2 binding conferred only a small decrease in lacZ reporter mRNA levels , which was limited to one genetic background ( Fig 5C ) , but a marked reduction in abundance of GFP mRNA ( Figs 1 and 2 ) . We considered that these different responses of the GFP and lacZ reporters to Scd6-MS2-F tethering might arise from differences in codon optimality . However , the sTAI values for the GFP and lacZ coding sequences , 0 . 37 and 0 . 30 , respectively , are both within one standard deviation of the mean sTAI value for all yeast genes of 0 . 35 [10] . Given the negative correlation between sTAI values and change in mRNA abundance conferred by dhh1Δ for native mRNAs [10] ( S11A Fig ) , it might be expected that lacZ mRNA ( in the absence of tethered Scd6-MS2-F ) would show increased abundance in dhh1Δ vs . WT cells; however we observed the opposite ( S5F Fig ) , and we made qualitatively similar findings for a heterologous GAL1-lacZ mRNA ( S5D Fig ) . Thus , these lacZ mRNAs behave more like mRNAs with optimized codons , except that the magnitudes of their reductions in dhh1Δ cells ( 2 . 5 to 4 . 5-fold ) exceed the typical response of 10–20% reduced abundance seen for native codon-optimal mRNAs [10] . In addition , one might expect that tethering Scd6-MS2 would evoke greater Dhh1-mediated reduction in mRNA abundance for the less codon-optimal lacZ versus more codon-optimal GFP mRNA [10] , but again we found the opposite result . Hence , it is unlikely that differences in codon optimality underlie the different responses of these two reporters to tethered Scd6-MS2 . Finally , it is noteworthy that most of the 53 mRNAs exhibiting the largest TE increases in scd6Δ cells do not exhibit increases in mRNA abundance ( S7F Fig ) , indicating that Scd6 frequently decreases TE without reducing mRNA abundance of native mRNAs—as we observed for the lacZ reporter on tethering Scd6-MS2 . On the other hand , a proportion of the 53 mRNAs do exhibit increased mRNA abundance in parallel with increased TE in scd6Δ cells ( found in upper quartile of box plot in col . 1 of S7F Fig ) —implying coupled repression of TE and mRNA abundance by Scd6—as we observed for the GFP reporter on tethering Scd6-MS2 . Another unexpected finding from the tethering assays was that reductions in reporter mRNA levels on tethering Scd6-MS-F are enhanced in dhh1Δ cells , increasing the repression ratio of GFP mRNA abundance ( Fig 2C and 2D ) and uncovering a repression of lacZ reporter mRNA abundance that was barely detectable in WT cells ( Fig 5C and 5D ) . These observations might indicate that Dhh1 interferes with mRNA degradation evoked by tethered Scd6-MS2-F . This influence of Dhh1 was not seen for the five native mRNAs presented as exemplars of Scd6 translational repression ( Fig 8C and S8A–S8D Fig ) , which all exhibit higher rather than lower mRNA levels in dhh1Δ cells . Moreover , increased mRNA levels in dhh1Δ cells holds for a large proportion of the group of 53 mRNAs whose TE was most strongly derepressed in the scd6Δ mutant ( S7F Fig , col . 3 ) . However , there is also a fraction of these mRNAs that do resemble the reporter mRNAs on Scd6-MS2 tethering in showing decreased mRNA abundance in dhh1Δ cells ( S7F Fig , bottom quartile in col . 3 ) . More work is required to understand the differing responses to Dhh1 for different Scd6 targets . In summary , our results , in combination with previous findings on yeast Scd6 [22] , support a model wherein recruitment of Scd6 to an mRNA , directed by or stabilized by its interaction with eIF4G , enables Scd6 to recruit other effectors of mRNA decapping/degradation and translational repression including , but not limited to , Dcp1/Dcp2 and Dhh1 , and possibly also to interfere directly with recruitment of the 43S PIC by binding to the C-terminus of eIF4G ( Fig 9 , ( i-iii ) ) . Decapping by Dcp1/Dcp2 and subsequent degradation of the mRNA can proceed concurrently with translational repression ( Fig 9 , ( iv ) ) . Based on our findings that ccr4Δ enhances mRNA turnover and translational repression , we suggest that association of the Ccr4/Not complex with the mRNA , or deadenylation of the mRNA , interferes with the ability of Scd6 to associate with the mRNA or recruit decapping activators and translational repressors , thereby diminishing Scd6-enhanced mRNA degradation and translational repression in WT versus ccr4Δ cells . Further work will be required to determine whether Scd6 is recruited to specific mRNAs by 3’UTR sequences or RNA binding proteins unique to its mRNA targets , or whether intrinsic features of mRNAs ( sequences or other binding proteins ) confer a heightened sensitivity to Scd6 that would be recruited broadly to most mRNAs by eIF4G . Our identification of native mRNAs targeted by Scd6 for translational repression sets the stage for efforts to reconstitute the repressive function of Scd6 and its associated decapping activators in a purified translation system , and thereby elucidate their molecular mechanisms of translational control .
Plasmids employed in this study are listed in Table 1 . Plasmids containing constructs encoding FLAG-tagged MS2-CP fusions to Npl3 , Sbp1 , and Scd6 ( S13 Fig , left ) were constructed by first PCR-amplifying NPL3 , SBP1 or SCD6 respectively with their native endogenous promoter ( ~500 bp upstream flanking sequence ) plus their coding sequence minus the stop codon , from genomic DNA of WT strain BY4741 , with primers containing a gene-specific restriction site at the N-terminus , and an XhoI site and XbaI/SpeI site at the C-terminus . The following primers were used: ( i ) NPL3 ( forward primer with SpeI site , 5’-ACGAGGACTAGTTATCAATATGCAAATGCTCGGC-3’; reverse primer with XhoI and SpeI sites , 5’-ACGAGCACTAGTCTCGAGCCTGGTTGGTGATCTTTCACG ) ; ( ii ) SBP1 ( forward primer with XbaI site , 5’-ACGAGCTCTAGATCATCGAGCGGAAAATATTG-3’; reverse primer with XhoI and XbaI sites , 5’-ACGAGCTCTAGACTCGAGTTCTTGCTTTTCTTCAGAACC-3’ ) ; ( iii ) SCD6 ( forward primer with SpeI site , 5’-ACGAGGACTAGTTGCTCGTAACAATCTTGG-3’; reverse primer with XhoI and SpeI sites , 5’-ACGAGGACTAGTCTCGAGAAATTCAACGTTGGAAGGAGG-3’ ) . The amplified fragments were inserted between the XbaI/SpeI sites of YCplac111 to generate YCplac111-NPL3 , YCpLac111-SBP1 , or YCpLac111-SCD6 , respectively . The MS2-CP CDS was PCR-amplified from plasmid pJC236 with primers containing an XhoI site and encoding a flexible linker ( Gly-Gly-Gly-Gly-Gly-Ser ) at the N-terminus , 3xFLAG epitopes , a stop codon and an overlapping sequence ( for fusion PCR ) at the C-terminus , using forward primer 5’-ATTCATCTCGAGGGTGGTGGTGGTGGTTCTGCTTCTAACTTTACTCAGTTCGTT-3’ and reverse primer 5’-TTACTTGTCATCGTCATCCTTGTAGTCGATGTCATGATCTTTATAATCACCGTCATGGTCTTTGTAGTCGTAGATGCCGGAGTTTGCTGCGAT-3’ . Next , the 3’UTR from each gene was amplified from genomic DNA of BY4741 , with primers containing an upstream overlapping sequence ( for fusion PCR ) and a downstream XmaI site , using primers: ( i ) NPL3 3’UTR , forward primer 5’-CATGACATCGACTACAAGGATGACGATGACAAGTAAGCCATTTATATAGTTGAGAAAAAA-3’; reverse primer 5’-ATTTATCCCGGGTACCTATTCTGGCGTGTAATCCTTATCA-3’ ) ; ( ii ) SBP1 3’UTR , forward primer 5’-CATGACATCGACTACAAGGATGACGATGACAAGTAATTACTTCTTACCCACATCCCTATT-3’; reverse primer 5’-ATTTATCCCGGGTACCTCTCCGAGGTAGTGAACCATTGAG-3’ ) ; and ( iii ) SCD6 3’UTR , forward primer 5’-CATGACATCGACTACAAGGATGACGATGACAAGTAAAATGATGTTTCTATGTAAATTAAGTA-3’; reverse primer 5’-ATTTATCCCGGGTACCCTTTTCTTGTAGTTTGTTGTTCTTAC-3’ ) . Fragments containing linker-MS2CP-FLAG-3’UTR sequences for each gene were generated by fusion PCR using the amplified fragments above , and inserted between the XhoI/XmaI sites of YCplac111-NPL3 , YCplac111-SBP1 , or YCplac111-SCD6 , to generate the constructs pQZ125 ( NPL3-MS2-F ) , pQZ126 ( SBP1-MS-F ) and pQZ127 ( SCD6-MS2-F ) . Plasmids encoding MS2-FLAG control proteins pQZ128 ( PNPL3-MS2-F ) , pQZ129 ( PSBP1-MS2-F ) and pQZ130 ( PSCD6-MS2–F ) ( S13 Fig , right ) were constructed by a strategy similar to that described above but with the NPL3 , SBP1 and SCD6 CDSs absent from the final constructs and an ATG added at the beginning of the MS2 CP-encoding fragment . The specific primers for these constructions were: ( i ) PNPL3-MS2-F: forward primer for NPL3 promoter with SphI site 5’-ACGAGGGCATGCTATCAATATGCAAATGCTCGGCTC-3’; reverse primer for NPL3 promoter with ATG 5’-CATTATCCTTATGGTTTTAGCGTAATT-3’; forward primer for MS2-NPL3 3’UTR with ATG 5’-AATTACGCTAAAACCATAAGGATAATGGGTGGTGGTGGTGGTTCTGCTTCT-3’; reverse primer for MS2-NPL3 3’UTR with KasI site 5’-ATTTATGGCGCCTATTCTGGCGTGTAATCCTTATCA-3’ ) ; ( ii ) PSBP1-MS2-F: forward primer for SBP1 promoter with SphI site 5’-ACGAGGGCATGCTCATCGAGCGGAAAATATTGAAAA-3’; reverse primer for SBP1 promoter with ATG 5’-CATATTTTTCTTCGTTTGAGGGTTTTC-3’; forward primer for MS2-SBP1 3’UTR with ATG 5’-GAAAACCCTCAAACGAAGAAAAATATGGGTGGTGGTGGTGGTTCTGCTTCT-3’; reverse primer for MS2-SBP1 3’UTR with XmaI site 5’-ATTTATCCCGGGTACCTCTCCGAGGTAGTGAACCATTGAG-3’ ) ; ( iii ) PSCD6-MS2-F: forward primer for SCD6 promoter with SphI site 5’-ACGAGGGCATGCTGCTCGTAACAATCTTGGCCTAGC-3’; reverse primer for SCD6 promoter with ATG 5’-CATTGCCTTGCTGCTGTTTTTCGATGA-3’; forward primer for MS2-SCD6 3’UTR with ATG 5’-TCATCGAAAAACAGCAGCAAGGCAATGGGTGGTGGTGGTGGTTCTGCTTCT-3’; reverse primer for MS2-SCD6 3’UTR with XmaI site 5’-ATTTATCCCGGGTACCCTTTTCTTGTAGTTTGTTGTTCTTAC-3’ ) . The GFP reporter plasmid pJC429 was described previously [6] . The lacZ reporter plasmid pQZ131 was generated by PCR-amplifying the lacZ CDS sequence from GCN4-lacZ reporter plasmid p180 , adding an SphI site and an ATG to the N-terminus ( forward primer 5’-AAACTTGCATGCTTACGGAT-3’ ) , and a PacI site to the C-terminus ( reverse primer 5’-ACGAGCTTAATTAATTTTTGACACC-3’ ) . The resulting fragment was inserted between the SphI/PacI sites of pJC429 . Plasmid pQZ145 ( DCP2 URA3 ) was generated by inserting into pRS316 a 4 . 3 kb XbaI-XmaI DCP2 fragment from pRS315-DCP2 . Plasmids pQZ139 ( ΔLSm-Scd6-MS2-F ) and pQZ142 ( ΔRGG-Scd6-MS2-F ) were generated by deleting the CDS for amino acids Q3-D78 or S287-N318 , respectively , of pQZ127 by site-directed mutagenesis ( GenScript USA Inc ) . All plasmids were screened by restriction digestion and DNA sequencing was conducted to verify the presence of the intended inserts . Yeast strains employed in this work are listed in Table 2 . Strain QZY126 was constructed by transforming HFY114 with a DNA fragment containing dhh1Δ::kanMX4 and including ~400 bp of sequences from both upstream and downstream of DHH1 that was PCR-amplified from genomic DNA of strain 3858 and selecting on YPD medium containing G418 . Strain QZY128 was constructed by transforming strain CFY1016 ( dcp2Δ::HIS3 ) harboring pQZ145 ( DCP2 URA3 ) with the dhh1Δ::kanMX4 cassette as above , and evicting pQZ145 by counter-selection growth on 5-FOA medium . FZY843 was constructed by transforming CFY1016 with a fragment containing scd6Δ::KanMX4 that was PCR-amplified from yeast strain 5544 . Strains SYY2352 and SYY2353 were constructed by transforming HFY114 with a DNA fragment harboring the scd6::KanMX6 null allele , which contains 400 bp from both upstream and downstream of SCD6 with the coding region replaced by a 1447-bp KanMX6 cassette . Gene disruptions were confirmed by PCR analysis of chromosomal DNA using the appropriate primers . Unless otherwise noted , all strains were grown at 30°C in synthetic complete medium without leucine or uracil ( SC-L-U ) with 2% galactose/2% raffinose replacing dextrose . All cultures were grown for at least two cell divisions and harvested at mid-log phase ( OD600 = 0 . 6–0 . 7 ) . For Western blot analysis , whole-cell extracts ( WCEs ) from at least three biological replicates ( independent transformants ) were prepared by trichloroacetic acid ( TCA ) extraction as previously described [43] . Aliquots of WCEs were resolved by 4–20% SDS-PAGE , transferred to PVDF membrane and probed with antibodies against GFP ( Covance ) , Prt1 [44] , FLAG epitope ( Sigma ) , or Gcd6 [45] . Immune complexes were detected using the Pierce enhanced chemiluminescence ( ECL ) system and autoradiography; and signal intensities were quantified by scanning densitometry using NIH ImageJ software . Assays of β-galactosidase activity in WCEs were performed as described previously [46] . At least four biological replicates ( and two technical replicates per transformant ) were employed for all β-galactosidase assays . For polysomes profiles , 300 mL of cells were treated with 50 μg/ml cycloheximide for 5 min prior to harvesting . WCEs were prepared in 1x breaking buffer ( 20 mM Tris-HCl , pH 7 . 5 , 50 mM KCl , 10 mM MgCl2 , 1 mM DTT , 50 μg/ml cycloheximide , 1 mM PMSF , Complete EDTA-free Protease Inhibitors , 1U/μl SUPERase-In RNase inhibitor ) by vortexing with glass beads , followed by two cycles of centrifugation for 10 min at 15 , 000 rpm at 4°C . 15 OD260 units of cleared lysate were loaded on 15%-45% ( w/w ) sucrose gradients prepared on a Biocomp Gradient Master in 1x breaking buffer and centrifuged at 41 , 000 rpm for 73 min at 4°C in a Beckman SW41Ti rotor . Gradients were fractionated with a Brandel Fractionation System and ribosomal peaks were visualized at 254 nm with an Isco UV detector . Fractions ( 0 . 7 mL ) were precipitated overnight at -20°C using 1 . 5 volumes RNA precipitation mix ( 95% EtOH , 5% NaOAc ) , and centrifuged at 15 , 000 rpm for 30 min to pellet RNA/protein . Pellets were washed in 1 mL cold 80% EtOH , dried in a speed vacuum , and resuspended in 100 μl AE buffer ( 50 mM NaOAc , 10 mM EDTA , pH 5 . 2 ) with 1% SDS . Fractions were extracted by adding 350 μl QIAzol lysis reagent and incubating 5 min at room temperature , adding 70 μL chloroform and incubating 2–3 min at room temperature , and centrifuging at 15 , 000 rpm for 15 min at 4°C . The aqueous phase ( ~300 μl ) was transferred to a new collection tube , thoroughly mixed with 1 volume of 100% EtOH , and applied to a purification column ( RNA Clean and Concentrator™ kit , Zymo Research ) to isolate RNA following the vendor’s instructions . RNA from each polysomal fraction was eluted with 25 μl RNase-free water , and 2 μl/reaction were used for first strand cDNA synthesis , as described below . Yeast cultures ( 25 mL ) were harvested by centrifugation , and the resulting pellets were resuspended in 400 μL AE buffer ( 50 mM NaOAc , 10 mM EDTA , pH 5 . 2 ) with 1% SDS . Cell suspensions were extracted twice with AE buffer-equilibrated phenol , twice with phenol/chloroform/isoamyl alcohol ( 25:24:1 ) , once with chloroform/isoamyl alcohol ( 24:1 ) , and then precipitated at -20°C with 2 volumes RNA precipitation mix ( 95% EtOH , 5% NaOAc ) . RNA pellets were recovered by centrifugation at 15 , 000 rpm for 30 min ( 4°C ) , washed once with 1 mL cold 80% EtOH , dried in a speed vacuum , and resuspended in 50 μL RNase-free water . Total RNA concentration was calculated at A260 using a NanoDrop spectrophotometer . For RT-qPCR , reverse transcription was carried out using SuperScript III First-Strand cDNA Synthesis SuperMix ( Invitrogen ) , with random primers and either 1 μg total cellular RNA from above or 2 μL polysomal RNA isolated from sucrose gradient fractions . qPCR was carried out using Brilliant III Ultra-Fast SYBR Green Master Mix ( Agilent ) in a Mx3000P System ( Stratagene ) and the following oligonucleotide pairs: GFP ( 5’-GGCTAGCAAAGGAGAAGAACTC-3’; 5’-CCGTATGTTGCATCACCTTC-3’ ) , lacZ ( 5’-ACCAACGTAACCTATCCCATTAC-3’; 5’-TTCCTGTAGCCAGCTTTCATC-3’ ) , ACT1 ( 5’-TGTGTAAAGCCGGTTTTGCC-3’; 5’-GATACCTCTCTTGGATTGAGCTTC-3’ ) . Levels of reporter mRNA relative to actin were calculated using the ΔCt method . For Northern blot analysis , 10 μg of total RNA/lane were denatured with glyoxal/DMSO , separated on 1 . 4% agarose gels , transferred to nylon membranes , and probed with [32P]-end-labeled DNA oligonucleotides complementary to GFP or yeast PYK1 transcripts . Blots were exposed to PhosphorImager screens and scanned using a Storm scanner . Measurements of GFP mRNA half-lives . A previously described protocol [47] was employed with the following modifications . Yeast transformants were cultured in 330 mL of SC-U , -L with 2% Galacose and 2% Raffinose to an A600 of ~0 . 8 . An aliquot of 30 mL , representing the “0 min” time point , was poured onto ice and collected by centrifugation at 3 , 000 rpm in a Beckman JS-4 . 2 rotor for 5 min at 4°C . The cell pellet was resuspended and transferred to a 1 . 5 mL Eppendorf tube , and spun at 4°C in a microfuge at top speed ( 15 , 000 rpm ) for 2 min . Residual medium was aspirated and cell pellets were immediately frozen on dry ice . The rest of the culture was collected by centrigugation at room temperature for 8 min at 7 , 000 rpm in an SLA3000 rotor . The cell pellet was resuspended in an equal volume of SC-U , -L 4% glucose medium pre-warmed to 30°C , and aliquots of 30 mL were harvested exactly as described for the “0 min” sample at 2- , 4- , 6- , 8- , 10- , 15- , -20 , -30 , -40 , and 60 min after glucose containing medium was added . Frozen cells pellets were stored at -80°C until being thawed for isolation of total RNA and qRT-PCR analysis , which was conducted as described above . Changes in reporter protein expression ( ΔGFP protein or Δβ-galactosidase ) or reporter mRNA abundance ( ΔGFP mRNA or ΔlacZ mRNA ) were calculated as ratios of mean values of protein or mRNA expression measured in 3 or more biological replicates of transformants expressing Scd6-MS2-F vs . MS2-F alone . The propagated S . E . M . s for the resulting mean ratios were computed as ( X/Y ) * ( √[ ( SEx/x ) 2+ ( SEy/y ) 2] ) , where X , SEx , and x are the mean , standard error of the mean , and highest values for reporter protein input , respectively; and Y , SEy , and y are the corresponding values for the mRNA input . Unpaired t-tests were performed to compare the mean changes in reporter protein or mRNA expression between wild type and mutant strains . Changes in TE of reporter mRNA on tethering Scd6-MS2 were determined by calculating the TE of the reporter , as the ratio of reporter protein to reporter mRNA expression , measured in 3 or more pairs of independent transformants expressing Scd6-MS2-F or MS2-F alone , from which the mean ΔTE and S . E . M . was calculated . Unpaired t-tests were performed to compare the resulting mean ΔTE values between wild type and mutant strains . Ribosome profiling and RNA-Seq analysis were conducted in parallel essentially as described previously [48] on isogenic strains in the W303 background HFY114 ( WT ) , CFY1016 ( dcp2Δ ) , SYY2353 ( scd6Δ ) , QZY126 ( dhh1Δ ) , FZY843 ( dcp2Δscd6Δ ) , and QZY128 ( dcp2Δdhh1Δ ) , providing two biological replicates of each genotype . Strains growing exponentially in YPD medium at 30°C were harvested by vacuum filtration at room temperature , and quick-frozen in liquid nitrogen . Cells were lysed in a freezer mill with lysis buffer ( 20 mM Tris [pH 8 . 0] , 140 mM KCl , 1 . 5 mM MgCl2 , 1% Triton , 500 μg/mL cycloheximide ) . For ribosome footprint library preparation , approximately 60 A260 units of extract were treated with RNAse I at 15 U per OD260 unit ( Ambion , #AM2295 ) for 1h at 25°C on a Thermomixer at 700 rpm , and 80S ribosomes were purified by sedimentation through a sucrose density gradient as described [49] . Ribosome-protected mRNA fragments ( footprints ) were purified by phenol and chloroform extractions [49] . Following size selection and dephosphorylation , a Universal miRNA cloning linker ( Synthesized by Integrated DNA Technologies ) was ligated to the 3’ ends of footprints , followed by reverse transcription , circular ligation , rRNA subtraction , PCR amplification of the cDNA library , and DNA sequencing with an Illumina HiSeq system . For RNA-Seq library preparation , total RNA was purified using miRNeasy Mini kit ( Qiagen ) from aliquots of the same extracts ( 30 OD260 units ) used for footprint library preparation , 5 μg total RNA was randomly fragmented at 70°C for 12 min in fragmentation reagent ( Ambion #AM8740 ) . Fragment size selection , library generation and sequencing were carried out as above , except Ribo-Zero Gold rRNA Removal Kit ( Yeast ) was employed to remove rRNAs after linker-ligation in lieu of poly ( A ) selection . As described [48] , linker sequences were trimmed from Illumina reads and the trimmed fasta sequences were aligned to the S . cerevisiae ribosomal database using Bowtie [50] . The non-rRNA reads ( unaligned reads ) were then mapped to the S . cerevisiae genome using TopHat [51] . Wiggle track normalization for viewing RPF or RNA reads in the IGV browser was conducted as follows . Wiggle files were produced from the alignment file , one each for genes on the Watson or Crick strand . The total reads on both strands were summed and a normalization factor q was calculated as 1 , 000 , 000 , 000/ ( total reads on W+C strands ) . Wiggle files were then regenerated by multiplying all reads by the factor q , yielding the number of reads per 1 , 000 million total reads ( rpkm ) . Statistical analysis of changes in mRNA , RPFs , or TE values between two replicates each of any two strains being compared was conducted using DESeq2 [52] , excluding any genes with less than 10 total mRNA reads in the 4 samples combined . RNA-Seq analysis of strains SYY2352 ( scd6Δ ) , SYY2353 ( scd6Δ ) , SYY2686 ( dhh1Δ ) , SYY2674 ( pat1Δ ) , and SYY2680 ( lsm1Δ ) was conducted as described previously [53] after culturing cells in YPD at 30°C; and the results will be described in full in a future publication . For all notched box-plots , constructed using a web-based tool at http://shiny . chemgrid . org/boxplotr/ , the upper and lower boxes contain the 2nd and 3rd quartiles and the band gives the median . If the notches in two plots do not overlap , there is roughly 95% confidence that their medians are different . RNA-Seq data employed in the analysis of mRNA changes in scd6Δ , dhh1Δ , pat1Δ , and lsm1Δ strains for the group of 83 mRNAs derepressed in scd6Δ cells have been deposited in the NCBI Gene Expression Omnibus ( GEO; http://www . ncbi . nlm . nih . gov/geo/ ) under the accessions numbers GEO:GSE107841 and GEO:GSE114428 . All other RNA-Seq or Ribo-Seq data generated in this study are deposited separately in GEO:GSE114892 .
|
Previous work showed that Scd6 homologs in Drosophila , Xenopus , and humans are associated with translationally repressed mRNAs in RNA granules , and that they interact with other mRNA silencing factors , including homologs of RNA helicase Dhh1 . However , there is little evidence that such Scd6 homologs are critical for translational repression or degradation of specific mRNAs in vivo . Yeast Scd6 interacts with the mRNA decapping enzyme and was shown to inhibit translation in cell extracts through binding to cap-binding translation initiation factor eIF4G . However , it was unknown whether Scd6 represses translation or accelerates degradation of any specific mRNAs in vivo , or whether Scd6 requires the decapping enzyme or its activators for such events . Here we show that tethering Scd6 stimulates the degradation or translational repression of reporter mRNAs in yeast , collaborating with Dhh1 for translational repression and the decapping enzyme Dcp2/Dcp1 for mRNA turnover . Using ribosome profiling we further identify groups of native mRNAs that appear to be targeted for degradation or translational repression by Scd6 and Dhh1 and find unexpectedly that Dcp2 is necessary for translational repression as well as enhanced degradation of most such mRNAs . Thus , Scd6 partners with Dhh1 and Dcp2 to mediate translational repression and degradation of specific mRNAs in vivo .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"statistics",
"messenger",
"rna",
"plasmid",
"construction",
"organisms",
"protein",
"expression",
"fungi",
"mathematics",
"dna",
"construction",
"molecular",
"biology",
"techniques",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"translation",
"initiation",
"gene",
"expression",
"molecular",
"biology",
"ribosomes",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"yeast",
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"rna",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"protein",
"translation",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"statistical",
"data"
] |
2018
|
Conserved mRNA-granule component Scd6 targets Dhh1 to repress translation initiation and activates Dcp2-mediated mRNA decay in vivo
|
The spread of disease through human populations is complex . The characteristics of disease propagation evolve with time , as a result of a multitude of environmental and anthropic factors , this non-stationarity is a key factor in this huge complexity . In the absence of appropriate external data sources , to correctly describe the disease propagation , we explore a flexible approach , based on stochastic models for the disease dynamics , and on diffusion processes for the parameter dynamics . Using such a diffusion process has the advantage of not requiring a specific mathematical function for the parameter dynamics . Coupled with particle MCMC , this approach allows us to reconstruct the time evolution of some key parameters ( average transmission rate for instance ) . Thus , by capturing the time-varying nature of the different mechanisms involved in disease propagation , the epidemic can be described . Firstly we demonstrate the efficiency of this methodology on a toy model , where the parameters and the observation process are known . Applied then to real datasets , our methodology is able , based solely on simple stochastic models , to reconstruct complex epidemics , such as flu or dengue , over long time periods . Hence we demonstrate that time-varying parameters can improve the accuracy of model performances , and we suggest that our methodology can be used as a first step towards a better understanding of a complex epidemic , in situation where data is limited and/or uncertain .
Our approach is based on three main components: an epidemiological model embedded in a state-space framework , a diffusion process for each time-varying parameter and an up-to-date Bayesian inference technique based on adaptive PMCMC . The main advantage of the state-space framework is the use of two sets of equations , the first set describes the propagation of the disease in the population and the second is for the observation process . This allows for consideration of unknowns and uncertainty both in the epidemiological mechanisms and in the partial observation of the disease: {x˙ ( t ) =g ( t , x ( t ) , θ' ( t ) , u ( t ) ) y ( t ) |x ( t ) ∼f ( h ( x ( t ) ) , θ' ( t ) ) ( 1 ) The first equation is for the epidemiological model , with x ( t ) representing the state variables ( for instance , S ( t ) the susceptibles , I ( t ) the infectious and R ( t ) the removed for the classical SIR model ) and θ' ( t ) the epidemiological parameters . The second is the observational process defined by probabilistic law f and a reporting rate on transformation of some state variable h ( x ( t ) ) because we may not be able to directly measure all state variables but just some or a function of them . In these equations , y ( t ) are partial observations of x ( t ) , u ( t ) is the process noise describing different form of stochasticity and the observational noise is included in f . In our applications , h ( x ( t ) ) will be the cumulative sum of new cases over the observation time step , that is generally the quantity observed by Public Health systems . Considering the time-varying parameters θ ( t ) as a subset of θ' ( t ) , we make the assumption that they evolve more or less randomly and do not follow a defined mathematical function . In the absence of prior information the use of diffusion motion allows us to impose few restrictions on the evolution of θ ( t ) . We consider that they follow a continuous diffusion process ( a discrete diffusion process was used in [35] ) : dθ ( t ) =σdB ( t ) ordlog ( θ ( t ) ) =σdB ( t ) ( 2 ) where σ is the volatility of the Brownian process ( dB ( t ) ) and will be estimated during the fitting process . The use of a Brownian process can be viewed as a weak hypothesis for the imposed motion of θ ( t ) and the volatility σ being a regularized factor . Intuitively , the higher the values of σ the larger the changes in θ ( t ) . The logarithm transformation avoids negative values which have no biological meaning . When prior knowledge on θ ( t ) is available this Brownian process can be modified to account for a drift in ( 2 ) ( see [36] ) . For the time-varying parameter , we focus on the parameter of the force of infection classically defined as: λ ( t ) =β ( t ) . S ( t ) . I ( t ) N ( 3 ) with β ( t ) the transmission rate usually defined by a sinusoidal function . The control or the behavior modification can also be taken into account: λ ( t ) =β ( t ) . ( S ( t ) εS ( t ) ) . ( I ( t ) εI ( t ) ) N ( 4 ) εi ( t ) describe the clustering of the population [39 , 40] but can also describe a reduction in the population due to voluntary avoidance behavior or social distancing . However due to the absence of structural identifiability properties [41 , 42] it should be very difficult to estimate simultaneously both β ( t ) and εi ( t ) . For model estimation we use Bayesian methods , coupling particle filter and MCMC for partially observed stochastic non-linear systems [36 , 43] ( see Methods ) . The implementation provided in SSM software [36] is used .
We start our demonstration by showing that it is possible to reconstruct both the trajectory of a SIRS model ( SIRS stands for Susceptibles , Infectious , Removed and Susceptibles again ) and that of the sinusoidal transmission rate . In this example , the trajectory of each variable has been simulated with a model for which all the parameters were known . Moreover we also knew the observation process that has generated the data , a Poisson law for the incidence with an observation rate equal to 1 . Fig 1 displays the reconstructed trajectories of both the incidence and the transmission rate highlighting the potential of the method . The parameter estimations are in perfect agreement with the values used to generate the observations and the estimation process has correctly converged ( S1 and S2 Figs ) . This clearly demonstrates the feasibility of accurately ascertaining the time evolution of the transmission rate and correctly estimating the Reff ( see Fig 2 ) . It is worth emphasizing that the SIRS model is a complicated example for different reasons . First , even with a constant transmission rate the SIRS model can generate oscillations ( damped oscillations , see [17 , 44] ) . Secondly , the model trajectories are not very sensitive , a modification of ± 10% can induce minor modifications of the trajectories that are inside or near the 95% CI of our inferences ( Fig 2 ) . Moreover in this example we have used initial conditions outside the attractor of the dynamics to generate transients that appear more realistic for real applications , but are more complex to reconstruct . The robustness of our approach has also been tested: ( i ) using long time series and initial conditions near the attractor ( Fig 3A and S3A Fig ) ; ( ii ) modifying the number of inferring parameters ( S4–S6 Figs ) , for instance estimating just the volatility parameter ( S7 and S8 Figs ) ; ( iii ) considering the possibility of not using the transformation log in the diffusion process ( S9 and S10 Figs ) and ( iv ) using a true β ( t ) with 2 or 3 periodic components ( Fig 3B and 3C and S3B and S3C Fig ) . We have also explored the performance of our approach by comparing their inferences to those of the true model . The re-estimation of the true model on its own data is displayed in S11–S13 Figs . Table 1 presents indices of the goodness-of-fit of the true model and models with time-varying β ( t ) with different number of parameters inferred . As expected , the error on β ( t ) is smaller when the true equation is used ( Table 1 ) . However , regarding the estimated incidence , the true model and our approach give similar results both in terms of mean and variance ( Table 1 ) . It could be argued that the price of the flexibility of our approach is a greater variability in some of the trajectory estimations ( Table 1 ) . Nevertheless the average dynamics are always estimated correctly . As misspecification is an important problem ( e . g . [45] ) we have also compared the performance of our approach to those of a misspecified seasonal SIRS model . We have thus used the example of a sinusoidal β ( t ) with two periodic components ( see Fig 3B ) and computed the indices of the goodness-of-fit of the true model with the SIRS model with 1 year sinusoidal β ( t ) and with our time-varying periodic β ( t ) . The results clearly show that our approach performed better than the misspecified model for the three trajectories analyzed , Incidence , β and Reff ( Table 2 ) . Once again the price of the flexibility of our approach is a greater variability in some of the trajectory estimations . However this is preferable to a large error in the median trajectories as occurred in those observed with the misspecified model ( Table 2 ) . Our methodology is also applicable to other more complex or simpler tasks . For instance , it can follow the time evolution of a parameter describing the availability of susceptibles , εS ( t ) ( Fig 4 and S14 and S15 Figs ) . Fig 4 shows the accurate reconstruction of the trajectory of the incidence and also of the trajectory of εS ( t ) that shifted at a given time point and decreased slightly thereafter . This highlights once again the potential of our approach as it is never easy to estimate a discontinuous dynamic with a continuous process ( 2 ) . In previous works , the dynamics of influenza in Israel have been analyzed using a discrete deterministic SIRS model and weekly data from Israel’s Maccabi health maintenance organization [20 , 46] . To describe the seasonality of this recurrent epidemic , the authors used a linear model between the transmission rate and local climatic variables , daily temperature and relative humidity [20 , 46] . We have re-analyzed their dataset ( but limited to 1998–2003 due to a modification in the reporting ) to reconstruct the time evolution of β ( t ) . Our results ( Fig 5 and S16 Fig ) clearly show the potential of our method , highlighting that the β ( t ) fluctuations are more irregular and complex than a simple sinusoidal function . Our last example is on dengue in Cambodia . Again the idea is to relax the assumption of a sinusoidal β ( t ) in a SEIR model . Monthly data from the capital Phnom Penh [47] , for which the meteorological data is available from the international airport , was used . We can accurately describe the 12 year time series and reconstruct the time evolution of β ( t ) ( Fig 6 and S17 Fig ) . Our results stress that the β ( t ) oscillations are more complex than a simple sinusoidal function . Sometimes bi-modality occurs over one season . In general one observes a fast growth of β ( t ) and a slow decrease . Moreover the amplitude of the β ( t ) varies from year to year , perhaps depending on the fluctuations in the mosquito population and in the environment . Interestingly the peak in β ( t ) appears 1 to 2 months before the incidence peak . This delay can be explained by the extrinsic incubation period and might be used in a warning system . To explain the β ( t ) oscillations we have explored the potential effects of local and global climatic variables using wavelet decomposition [48] as one of our main underlying hypotheses is non-stationarity . We observed very significant coherency between β ( t ) and climate for the local climate for the seasonal mode ( Fig 7 and S18–S20 Figs ) and also for the 2–3 year components with global climatic variable ( S21 Fig ) . Thus , the rhythm of β ( t ) can be explained perfectly by climatic factors . Nevertheless , again mainly due to large non-stationarity , by using solely one or two climatic variables we are able to correctly describe dengue evolution in the short-term ( Fig 7C , red area ) but not over a large time period ( Fig 7C , blue area ) . This reflects the complexity of such a disease where the ecology of the vectors , the environment , the climate , the immune status of the human population and its behavior are all involved . This large non-stationarity association between dengue and climatic factors has recently been demonstrated using statistical models ( dynamic generalized linear models ) and data from a medium-sized city in Colombia [49] . The authors showed that dengue cases correlate with climatic variables ( temperature , rainfall , solar radiation and relative humidity ) but these correlations change over time , some intervals showing a positive association , while in others the association is negative [49] . The non-stationarity association between dengue and climate may be explained by the fact that a climatic variable has different effects depending on the biological cycle of the pathogen or of the vector . Moreover the effects of one climatic variable can also depend on other climatic variables potentially enhancing the non-stationarity association .
As there remain numerous uncertainties during the course of each epidemic , we are increasingly aware of the importance of developing adequate statistical and mathematical tools . Such tools need to take account of the time-varying nature of the underlying ecological and biological mechanisms as well as social and behavioral influences involved in an epidemic . Because of this , time-varying parameters modeled with a diffusion process , that track epidemiological patterns and update the key parameters according to data appear to be a worthwhile approach . Indeed developing a more complex model would be difficult considering the relative paucity of available data . We propose a flexible modeling framework that encompasses time-varying aspects of the epidemic . It does this via diffusion process equations for time-varying parameters and also considers uncertainty associated with key parameters and data . This data-driven framework for time-varying parameters has been coupled with simple stochastic models and a robust Bayesian procedure for inference . To test its efficiency , our proposed methodology was first applied to a toy model and then to real epidemiological examples . Our results clearly demonstrate the potential of our framework . Firstly , our methodology was able to accurately reconstruct both the incidence and the sinusoidal transmission rate of a simple SIRS model just based on noisy observations ( Figs 1–4 and S4 , S5 , S7 , S9 and S14 Figs ) . Based on these reconstructions one can also closely estimate Reff which is one of the key relevant epidemiological parameters . Our results also highlight the flexibility of our developed methodology . It can reconstruct the time evolution of a shifting parameter ( εS ( t ) , see Fig 4 and S14 Fig ) as well as an oscillating parameter that influences the nonlinear part of the model ( β ( t ) , see Figs 1–3 and S4 , S5 , S7 and S9 Figs ) . The comparison using goodness-of-fit indices with the inferred true model allows us to highlight the fact that our methodology performs as well for the observed incidence . Its flexibility results in greater variability in some other trajectories mainly β ( t ) and Reff ( Table 1 ) . Moreover , in the absence of knowledge of the true evolution of the transmission rate , our approach appears to capture the dynamic observed more accurately than a misspecified model ( Table 2 ) . Secondly , applied to real datasets , our framework is able , based solely on simple stochastic models , to reconstruct complex epidemics such as flu or dengue over long time periods ( Figs 5 and 6 ) . In such cases , the reconstruction of the time evolution of the transmission rate clearly stresses that , on real datasets , it is difficult to assimilate the dynamic of this parameter as a simple sinusoidal function . It is more irregular in amplitude and sometimes multi-modal over one season . Considering the paucity of information available regarding the complexity of the mechanisms involved during an epidemic , describing and fitting a full model for a given transmissible disease is always challenging . Our data-driven methodology can be used as a first step towards a better understanding of a complex epidemic , where data is limited or lacks certainty . Indeed most of the unknowns and uncertainties can be put into time-varying parameters . The potential effects of all these uncertainties can then be explored by analyzing the reconstructed time evolution of the time-varying parameters . See Fig 7 for such preliminary analysis of dengue in Phnom Penh . This allows a more thorough analysis of the influences and the interactions between both the human behavior and complex environmental drivers . In a recent paper [50] , the authors reviewed evidence of interactions between seasonal influenza virus and other pathogens ( bacteria or virus ) . They concluded that it is important to incorporate these different coinfecting pathogens in models of seasonal flu in order to get a better estimate of the burden of influenza . Our framework could be an alternative to the development of complex models with all the potential interactions between pathogens and to estimate the strength of the interactions . After reconstructing the time evolution of the transmission rate the statistical association between the coinfecting pathogens and the transmission rate could be tested . This screening may facilitate the construction of more complex models that could incorporate only the most significant coinfecting pathogens in the seasonal flu model . Our methodology also has other advantages . Taking account of the simplicity of the model used , and the fact that weak hypotheses on the dynamics of the time-varying parameters have been included , our proposed methodology can retrospectively test the impact of interventions . This has previously been done in the case of HIV epidemics [34–35] , where it was hypothesized that the reduction in the transmissibility was due to a modification of the sexual behavior in the population and the increase in the seropositive period duration due to the introduction of the first antiviral treatments . Evaluation of interventions has also been done recently in the case of the Ebola epidemic in West Africa [51] . The relative simplicity of our methodology is also suitable for short-term predictions and it can then easily be used to predict an epidemic in real time . Starting with a given estimated state defining the system , the fitting process can be run again each time new data is available and the new posteriors are used for new predictions [36] . This can inform public health decisions and indeed has been done recently to great effect in the case of the Ebola epidemics in West Africa [52] . A major challenge in model fitting is the reliability of data collected and also the non-identifiability of the mechanistic models that always have very rich dynamical behavior . The question of identifiability is too often avoided in epidemiological models applied to a topical Public Health issues . There is , however , considerable literature on this subject ( e . g . [41 , 42 , 53–55] ) . Identifiability is not evident even for a simple seasonal SIR model [56] . To solve this problem one can fit a combination of parameters or fix some of them ( the population size for instance ) [57] . In our applications there is a clear limitation due to practical non-identifiability of reporting rate and initial conditions . To fix these problems we have used informative priors ( see Method ) . Using informative priors or fixing some parameters gives very similar results ( compare Fig 1 and S4–S7 Figs ) . Related to this is the misspecification of models [45] . In our cases , as with other semi-mechanistic models the time-varying parameter methodology captures some of the information in the data but not in the mechanistic part of the model . If the model is misspecified due to lack of precision , it compensates for it and the dynamics of β ( t ) will drive improvements in the model to make it more complex and realistic ( Table 2 ) . If the model is misspecified to the extent that it creates mechanisms that do not exist , the reconstructed β ( t ) would compensate for these effects but it will be harder to interpret . In this work we have used simple mechanistic models . The proposed methodology is not limited to simple models . For instance , a two-strain dengue model has also been tested . In this case the main problem was linked to the unavailability of both seroprevalence and incidence for each strain . Indeed , one of the major difficulties with these multi-strain models is the identification of the initial conditions ( e . g . [58] ) . Nevertheless it is worth emphasizing that the Bayesian inference method used in our framework , PMCMC , the approximation of the likelihood is limited for a large number of parameters and/or equations [59] . In such cases testing other methodologies like ABC [60 , 61] is advisable . It is always difficult to fit complex models with rich behaviors based on very limited information . In this regard we agree with Metcalf et al . [62] who stressed that nowadays we need seroprevalence studies to quantify the immunological status of the population , because in most cases the magnitude of the outbreak is difficult to evaluate without precise seroprevalence data . To tackle the uncertainty and the non-stationarity of epidemics , our methodology , although it appears non-standard , makes important progress towards a better understanding of the mechanisms responsible for disease propagation . We believe that , should it form part of the development of the next predictive tools for Public Health , it will make a significant contribution to improving the understanding and control of infectious diseases in our increasingly uncertain world .
Among the various approaches developed to study nonstationary data , wavelet analysis is probably the most efficient . In particular , this method gives us the possibility of investigating and quantifying the evolution in time of the periodic components of a time series ( see [69] ) . Wavelets constitute a family of functions derived from a single function , the ‘‘mother wavelet” , Ψa , τ ( t ) , that can be expressed as the function of two parameters , one for the time position τ , and the other for the scale of the wavelets a , related to the frequency . More explicitly , wavelets are defined as: Ψa , τ ( t ) =1aψ ( t−τa ) The wavelet transform of a time series x ( t ) with respect to a chosen mother wavelet is performed as follows: Wx ( a , τ ) =1a . ∫−∞+∞x ( t ) . Ψ* ( t−τa ) . dt=∫−∞+∞x ( t ) . Ψa , τ* . dt where * denotes the complex conjugate form . The wavelet transform Wx ( a , τ ) represents the contribution of the scale a to the signal at different time positions τ . The computation of the wavelet transform is done along the signal x ( t ) simply by increasing the parameter τ over a range of scales a until all coherent structures within the signal can be identified . Here , as mother wavelet , we have used the Morlet wavelet [69] . With the wavelet approach , we can estimate the repartition of variance at different scale a and different time location τ . This is known as the wavelet power spectrum: Sx ( a , τ ) = | Wx ( a , τ ) |2 . An important point with the continuous wavelet is that the relationship between the wavelet frequency f0 and the wavelet scale a can be derived analytically . For the Morlet wavelet this relationship is given by: 1f=4πaf0+2+f02 Then when f0 = 2π , the wavelet scale a is inversely related to the frequency , f ≈ 1/a . This greatly simplifies the interpretation of the wavelet analysis and one can replace , on all equations , the scale a by the frequency f or the period 1/f . To determine the statistical relationship between two time series , wavelet coherence can be computed ( e . g . [48 , 70] ) : Rx , y ( f , τ ) = ( |〈Wx , y ( f , τ ) 〉|2|〈Wx ( f , τ ) 〉|2 . |〈Wy ( f , τ ) 〉|2 ) 1/2 where the angle brackets around terms indicate smoothing in both time and frequency , Wx ( f , τ ) is the wavelet transform of series x ( t ) , Wy ( f , τ ) is the wavelet transform of series y ( t ) , and Wx , y ( f , τ ) is the cross-wavelet spectrum . The values of wavelet coherence are between 0 < Rx , y ( f , τ ) < 1 . The wavelet coherency is equal to 1 when there is a perfect linear relation at particular time and scale between the two signals , and equal to 0 if x ( t ) and y ( t ) are independent . To complement this , phase analysis can be used to characterise the association between signals ( e . g . [48 , 70] ) . The phase difference provides information on the sign of the relationship ( i . e . , in phase or out of phase ) and can be computed , for complex mother wavelet , with the wavelet transform Wx ( f , τ ) as: ϕx ( f , τ ) =tan−1Im ( Wx ( f , τ ) ) Re ( Wx ( f , τ ) ) Similarly with the cross-wavelet transform Wx , y ( f , τ ) the phase difference between the two time series can be computed: ϕx , y ( f , τ ) =tan−1Im ( Wx , y ( f , τ ) ) Re ( Wx , y ( f , τ ) )
|
As our world becomes more and more globalized , infectious disease poses an ever-increasing threat to human health . The multitude of environmental and behavioral factors , which account for the spread of infectious diseases , are ever-evolving and thus infectious diseases propagation is complex . In the face of this complexity , mathematical models offer valuable tools to study the dynamics of epidemic diseases . Developing adequate statistical and mathematical tools , that take account of the time-varying nature of the different mechanisms responsible for disease propagation , remains a major challenge . To take this increasingly important aspect into consideration , we propose a flexible methodology that encompasses time-varying aspects of the epidemic . It does this via diffusion process equations for time-varying parameters . Considering the relative paucity of available data , our principal assertion is that it is preferable to use this flexible framework with time-varying parameters , that tracks epidemiological patterns , and updates the key parameters according to data , than to use a more complex model .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"evolutionary",
"rate",
"simulation",
"and",
"modeling",
"immunity",
"infectious",
"disease",
"epidemiology",
"epidemiology",
"biology",
"and",
"life",
"sciences",
"immunology",
"evolutionary",
"biology",
"evolutionary",
"immunology",
"evolutionary",
"processes",
"epidemiological",
"methods",
"and",
"statistics",
"research",
"and",
"analysis",
"methods"
] |
2018
|
Accounting for non-stationarity in epidemiology by embedding time-varying parameters in stochastic models
|
Calcium has a pivotal role in biological functions , and serum calcium levels have been associated with numerous disorders of bone and mineral metabolism , as well as with cardiovascular mortality . Here we report results from a genome-wide association study of serum calcium , integrating data from four independent cohorts including a total of 12 , 865 individuals of European and Indian Asian descent . Our meta-analysis shows that serum calcium is associated with SNPs in or near the calcium-sensing receptor ( CASR ) gene on 3q13 . The top hit with a p-value of 6 . 3×10-37 is rs1801725 , a missense variant , explaining 1 . 26% of the variance in serum calcium . This SNP had the strongest association in individuals of European descent , while for individuals of Indian Asian descent the top hit was rs17251221 ( p = 1 . 1×10-21 ) , a SNP in strong linkage disequilibrium with rs1801725 . The strongest locus in CASR was shown to replicate in an independent Icelandic cohort of 4 , 126 individuals ( p = 1 . 02×10-4 ) . This genome-wide meta-analysis shows that common CASR variants modulate serum calcium levels in the adult general population , which confirms previous results in some candidate gene studies of the CASR locus . This study highlights the key role of CASR in calcium regulation .
Calcium is the most abundant mineral in the human body contributing approximately one kilogram to the average adult human body mass . Whereas 99% of calcium is stored in the skeleton and teeth , the remaining 1% circulates in the bloodstream and is involved in many physiological processes including its function as a universal cellular signaling molecule [1]–[2] . Calcium plays a key role in membrane potential , which is important for muscle contraction , heart rate regulation and generation of nerve impulses . Calcium also influences bone metabolism , ion transport and many other cellular processes [3] . Approximately 2/5 of calcium in the extracellular fluid is found in blood serum . The level of serum calcium is under tight hormonal control with a normal concentration of 2 . 15–2 . 55 mmol/L . Serum calcium is under strong genetic control , with twin studies showing that the variance in total calcium due to genetic effects is between 50% and 78% [4]–[5] . While skeletal calcium is important in numerous clinical disorders , in particular bone and mineral disorders , the clinical role of serum calcium is less clear . Several [6]–[7] ( but not all [8] ) studies indicated that elevated serum calcium levels are associated with an increased risk of cardiovascular disease . Patients with hyperparathyroidism , who suffer from chronic hypercalcemia , have a high prevalence of hypertension and increased cardiovascular mortality [9]–[11] . However , the mechanisms underlying the putative association of serum calcium with increased cardiovascular morbidity and mortality remain unclear . Rare monogenic forms of hypo- or hypercalcemia have been described , including disorders involving the calcium-sensing receptor ( CASR , locus 3q13 ) gene . Heterozygous and homozygous CASR mutations that inactivate CASR are responsible , respectively , for familial hypocalciuric hypercalcemia , type 1 ( also known as familial benign hypercalcemia ) ( OMIM #145980 ) [12]–[13] and neonatal severe hyperparathyroidism ( OMIM #239200 ) [13] . On the other hand , mutations that result in CASR activation lead to autosomal dominant hypocalcemia ( OMIM #146200 ) [14] . Mutations in many other genes have also been found to lead to disturbances of serum calcium levels ( Table 1 ) . In the present study , we report results obtained from meta-analysis of genome-wide associations of serum calcium levels from four cohorts with a total of 12 , 865 participants . We first describe the design of the study and its main finding , that variants in CASR give rise to the strongest signals associated with serum calcium levels in both European and Indian Asian populations . Our results confirm previous studies showing that mutations in CASR are associated with serum calcium levels in young healthy women [15]–[16] and extend this observation to men and women across a large spectrum of age . We show that CASR is a key player in the genetic regulation of serum calcium in men and women from the general adult population .
We performed a meta-analysis for genome-wide associations of serum calcium , determined by subtracting the estimated amount of calcium bound to albumin from the total serum calcium , to infer the amount of ionized calcium ( see Materials and Methods ) . Our study included four cohorts: ( i ) 5404 European individuals from the Cohorte Lausanne ( CoLaus ) [17]–[18] , ( ii ) 5548 European and Indian Asian individuals from the London Life Sciences Population ( LOLIPOP ) Study from West London UK [19]–[20] , ( iii ) 1196 European individuals from the InCHIANTI Study ( Tuscany , Italy ) [21] , and ( iv ) 717 individuals of European descent from the Baltimore Longitudinal Study of Aging ( BLSA ) study based in the Baltimore-Washington DC area [22] , totaling 12 , 865 participants ( see Table 2 for more detailed characteristics of each cohort ) . Genome-wide association scans were performed first independently for each cohort using linear regression and then the effect sizes from each cohort were meta-analyzed ( see Materials and Methods ) . Due to the possibility of population substructure obscuring effects of genetic variants , meta-analysis was performed separately for ( i ) combined European and Indian Asian cohorts ( N = 12 , 865 ) and restricted to cohorts of ( ii ) European ( N = 8 , 919 ) , and ( iii ) Indian Asian descent ( N = 3 , 947 ) . The meta-analyses yielded 100 SNPs from the combined cohorts , 70 SNPs when restricting to European cohorts and 22 SNPs restricting to Indian Asian cohorts that exceeded the genome-wide significance threshold of 5×10-8 ( Figure 1A–1C ) ( the full list is provided in Table S2A , S2B , S2C ) . All SNPs reaching statistical significance clustered around the CASR locus at 3q13 . The most significant SNP in the ( i ) combined and ( ii ) European meta-analyses was rs1801725 ( p = 6 . 29×10-37 , p = 2 . 58×10-18 , respectively ) and in the ( iii ) Indian Asian meta-analysis was rs17251221 ( p = 1 . 07×10-21 ) . These two SNPs are less than 11 kb apart and are in high linkage disequilibrium with each other ( r2 = 0 . 946 , 0 . 494 , 1 . 0 , 1 . 0 in HapMap CEU , CHB , JPT , YRI , respectively ) , and therefore most likely derive from the same association signal . We find that rs1801725 explains 1 . 26% of the variance in serum calcium , with the effect sizes and standard errors of the serum calcium increasing T allele in individual cohorts shown in Figure 2 and Table S3 . According to our additive model , each rs1801725 T allele increases log10 serum calcium ( in units mmol/L ) by 3 . 61×10-3 , equivalent to a multiplicative effect of 1 . 008 on serum calcium ( see also Table S2 ) . At an average serum calcium level of 2 . 25 mmol/L , each rs1801725 T allele yields an increase of 0 . 01874 mmol/L , or 21% of one standard deviation of serum calcium levels in a normal population . The regional pattern of association of SNPs around the CASR locus , and their linkage disequilibrium with rs1801725 , are shown in Figure 3 . Of note , rs1042636 , which has been associated with decreased serum calcium [23] , also achieved genome-wide significance with the G minor allele associated with decreased serum calcuim ( p = 4 . 96×10-9 ) . However , conditional on the rs1801725 locus , located 12 bps upstream , the rs1042636 p-value became 3 . 32×10−4 , indicating that the two loci share contributions to serum calcium levels . To confirm the rs1801725 signal , we analyzed the association pattern with serum calcium in a separate cohort . We used a subset of 4 , 126 Icelandic individuals from the deCODE study [24]–[25] with serum calcium measurements . We found the rs1801725 T allele to be strongly associated with increased serum calcium ( p = 1 . 02×10-4 ) , replicating the key meta-analysis result . While only the CASR locus reached nominal genome-wide significance for association with serum calcium , the top regions with p<10-5 are shown in Table 3 . These SNPs cover 12 regions , the significance of which is displayed across cohorts in Figure S3 . There were no SNPs in other candidate genes ( which have previously been shown to be involved in disorders associated with disturbed serum calcium levels ) that were associated with serum calcium at genome-wide significance . The most significant SNPs within 500 kb of the gene transcripts are shown in Table 1 . Considering the set of 18 , 611 distinct SNPs mapping to the set of serum calcium candidate genes excluding CASR , we find no significant association ( at significance level 0 . 05 and applying the Bonferroni correction for multiple testing , giving a cut-off p-value of 2 . 69×10-6 , see also Figure S4 ) . Indeed , fixing the sample size and genome-wide significance threshold our study is well-powered ( ≥0 . 80 ) to detect SNPs explaining at least 0 . 31% of the variance . Therefore the common SNPs within the candidate genes ( excluding CASR ) likely play at best a small role in serum calcium regulation . We analyzed the association of the top SNP with several calcium-related outcomes ( coronary heart disease , myocardial infarction , hypertension , stroke , osteoarthritis , osteoporosis and kidney stones ) . The number of cases and controls for each outcome and each cohort is given in Table S4 . Logistic regression including age and pseudosex ( see Materials and Methods ) as covariates did not find any significant association between rs1801725 and the calcium-related outcomes , after correcting for multiple testing ( effect sizes and standard errors for the T allele are listed in Table S5 ) . Power calculations show that given the sample sizes for the clinical traits above , our study has good power ( ≥0 . 80 ) to detect odds ratios of 1 . 20 , 1 . 13 , 1 . 77 , 1 . 27 , 1 . 27 , 1 . 24 and 1 . 75 , respectively . As the smallest p-values from calcium-related traits were for osteoarthritis and osteoporosis ( bonferroni-corrected p = 0 . 21 , 0 . 44 , respectively ) , we further investigated bone density traits . None of deCODE hip bone mineral density or spine bone mineral density ( N = 6657 and 6838 , respectively ) nor InCHIANTI total bone density , trabecular bone density , cortical bone density , cortical bone thickness or cortical bone area ( N = 1196 ) bonferroni-adjusted p-values for eight traits were significant .
This genome-wide scan of 12 , 865 individuals revealed CASR as the most significant ( and only genome-wide significant ) locus influencing serum calcium levels . Specifically , we found evidence for a strong association of SNPs located in the CASR locus with serum calcium levels in both Europeans and Indian Asians . The strongest locus in CASR was further shown to replicate in an independent Icelandic cohort of 4 , 126 individuals . The top signal ( rs1801725 , 2956G>T ) explains 1 . 26% of the variance in serum calcium . Indeed , this is similar to results from other GWAS of human height [26]–[29] , body mass index [30]–[31] , serum urate [32]–[34] and serum lipid concentrations [34]–[36] , for which the genome-wide significant loci uncovered thus far collectively explain only a small fraction of the phenotypic variance ( usually at least one order of magnitude less than the total additive genetic variance estimated from heritability studies [37]–[38] ) . The rs1801725 T allele ( A986S ) was associated with higher serum calcium , consistent with previous findings ( see Table S6 ) . The rs1801725 polymorphism ( with T allele frequencies of 16 . 76% , 19 . 98% in European and Indian Asian cohorts , respectively ) affects serum calcium levels of a substantial proportion of the population . The rs1801725 polymorphism encodes a missense variant in exon 7 of the CASR gene , which leads to a non-conservative amino-acid change ( serine substitution for alanine-986 , A986S corresponding to nucleotides 2956G>T ) in the cytoplasmic tail of CASR . In vitro studies showed that mutations within the C-terminal tail may influence several aspects of CASR function , such as signal transduction , intracellular trafficking and cell surface expression [39]–[41] . However , PolyPhen predicts rs1801725 to be a benign substitution . It is presently unclear whether this substitution gives rise to functional variants , as functional studies have yielded conflicting results [42]–[43] . Deep sequencing of this region may help identifying the causal variants . While it is still not possible to infer a direct causal role , it is of interest to note that the SNP gives rise to an amino acid change in the C-terminal tail of CASR , a domain which plays a key role in the receptor function and may potentially influence intracellular trafficking following CASR activation by extracellular calcium . Several studies have reported associations of A986S and nearby CASR mutations with various phenotypes . The A986S CASR polymorphism has been associated with variations in circulating calcium levels in healthy adults in some studies [15] , [23] , [44]–[45] , but not in others [46]–[47] . The fact that the latter studies were underpowered ( sample size ranging from 148 to 1252 ) to detect a small effect size likely explains these inconsistent results . The rs1042636 ( R990G ) polymorphism has been associated with the magnitude of parathyroid hormone ( PTH ) secretion in patients with primary hyperparathyroidism [48] , and preliminary results suggest that it could influence response to cinacalcet , a calcimimetic used to treat secondary hyperparathyroidism in patients with end-stage renal disease [49] . In a meta-analysis , 986S was associated with a 49% increased risk ( P = 0 . 002 ) of primary hyperparathyroidism [47] , [50]–[51] . Among patients with primary hyperparathyroidism , the AGQ haplotype ( i . e . 986A , 990G , 1011Q , which is associated with lower serum calcium levels and hypercalciuria [52] ) was associated with increased risk , and the SRQ haplotype with decreased risk , of kidney stones [50] . CASR has been previously considered as a candidate gene for osteoporosis [53] and coronary heart disease as well as increased total and cardiovascular mortality [54] . In our meta-analysis , we found no significant association of rs1801725 with these calcium-related phenotypes . A recent meta-analysis focusing on effects of candidate genes on osteoporosis also reports negative results for CASR [55] . Furthermore , results on the association of elevated serum calcium with increased cardiovascular risk in the general population are controversial [6]–[8] . It is therefore not clear to what extent serum calcium might predict cardiovascular risk . The SNPs identified in this meta-analysis could serve as genetic instruments in future studies , such as Mendelian randomization analysis in longitudinal cohorts , to further investigate the causal effect of serum calcium on osteoporosis and on cardiovascular disease risk ( see Table S5 for rs1801725 effects and standard errors ) . Our meta-analysis suffers from some limitations . First , we used corrected serum calcium and not directly measured ionized serum calcium . The correlation between corrected serum calcium and ionized serum calcium varies between 0 . 66 and 0 . 87 [56]–[58] . We can hypothesize that the association of ionized serum calcium with CASR variants would be stronger than the one with corrected serum calcium because ionized calcium is the form physiologically active on CASR . Second , data on serum phosphate , PTH or vitamin D are not available , so that we cannot explore further these relationships . Third , sample sizes for calcium-related clinical traits were limited , many clinical traits in CoLaus were self-reported instead of clinically diagnosed , and we incur a multiple testing penalty due to the number of clinical traits posited to be associated with serum calcium . However , the major strengths of the study are the hypothesis-free nature of GWAS studies , the large sample meta-analysis and the inclusion of multiple populations .
CoLaus is a population-based sample from Lausanne , Switzerland , consisting of 5435 individuals between 35 and 75 years old ( after QC ) of which a subset of 5404 had available serum calcium measurements . The study design and protocols have been described previously [17]–[18] . The CoLaus study was approved by the Institutional Ethic's Committee of the University of Lausanne . The London Life Sciences Prospective Population Study ( LOLIPOP ) is an ongoing population-based cohort study of ∼30 , 000 Indian Asian and European white men and women , aged 35–75 years living in West London , United Kingdom [59] . All study participants gave written consent including for genetic studies . The LOLIPOP study is approved by the local Research Ethics Committee . The participants included in the present study are a subset of 3947 Indian Asians and 1601 Europeans from the LOLIPOP cohort study . LOLIPOP individuals are separated by origin as well as the genotyping platform , with IAA , IAI or IAP denoting Indian Asians genotyped on Affymetrix , Illumina or Perlegen platforms , respectively , and EWA and EWI denoting Europeans genotyped on Affymetrix or Illumina platforms , respectively ( see Table S1 ) . The InCHIANTI study is a population-based epidemiological study aimed at evaluating the factors that influence mobility in the older population living in the Chianti region in Tuscany , Italy . The details of the study have been previously reported [60] . Overnight fasted blood samples were taken for genomic DNA extraction , and measurement of serum calcium . For this study , 1196 subjects with serum calcium and GWAS data were analyzed . The study protocol was approved by the Italian National Institute of Research and Care of Aging Institutional Review and Medstar Research Institute ( Baltimore , MD ) . The Baltimore longitudinal study on Aging ( BLSA ) study is a population-based study aimed to evaluate contributors of healthy aging in the older population residing predominantly in the Baltimore-Washington DC area [61] . Starting in 1958 , participants are examined every one to four years depending on their age . Blood samples were collected for DNA extraction . This analysis focused on a subset of the participants ( N = 717 ) of European ancestry . The BLSA has continuing approval from the Institutional Review Board ( IRB ) of Medstar Research Institute . Approval was obtained from local ethic committees for all studies and all participants signed informed written consent . The deCODE study consists of individuals who visited a private outpatient laboratory , the Laboratory in Mjodd , Reykjavik , Iceland between 1997 and 2008 . The main referral center for this laboratory is a multispecialty medical clinic in Reykjavik ( Laeknasetrid ) . For the serum calcium analysis we used information on 4 , 126 individuals with both genome-wide SNP data and measured serum calcium and serum albumin . The samples for bone density analysis have previously been described in detail [24]–[25] . For this study 6 , 657 individuals with total hip bone mineral density ( BMD ) and 6 , 838 individuals with lumbar spine BMD and SNP data were available for analysis . All participants gave informed consent and the study was approved by the Data Protection Commission of Iceland ( DPC ) and the National Bioethics Committee of Iceland . For each CoLaus participant a venous blood sample was collected under fasting conditions . Measurements were conducted using a Modular P apparatus ( Roche Diagnostics , Switzerland ) . Total serum calcium was measured by O-cresolphtalein ( 2 . 1% – 1 . 5% maximum inter and intra-batch CVs ) ; albumin was measured by bromocresol green ( 2 . 5% – 0 . 4% ) . To further characterize the identified genetic variants , we analyzed the association with several outcomes postulated to be correlated with serum calcium . Within the CoLaus study , we have questionnaire responses to queries about personal histories of osteoporosis , osteoarthritis , myocardial infarction and stroke in addition to clinical data determining hypertension status , defined as previously described [17] . The assessment of LOLIPOP study participants was carried out by a trained research nurse , during a 45 minute appointment according to a standardized protocol and with regular QC audits . An interviewer-administered questionnaire was used to collect data on medical history , family history , current prescribed medication , and cardiovascular risk factors . Physical assessment included anthropometric measurements ( height , weight , waist , hip ) and blood pressure . Blood was collected after an 8 hour fast for biochemical analysis , including glucose , insulin , total and HDL cholesterol and triglycerides , and whole blood was taken for DNA extraction [59] . InCHIANTI serum albumin concentrations were determined as percentage of total protein using agarose electrophoretic technique ( Hydragel Protein ( E ) 15/30 , Sebia , Issy-les-Moulineaux , France ) . Serum calcium was measured using calorimetric assay ( Roch Diagnostic , GmbH , Mannheim , Germany ) by a Roche-Hitachi autoanalyzer ( The intra-assay CV and 0 . 9% and the inter-assay CV was 1 . 5% ) . Measures of bone density , bone dimensions and osteoporosis diagnosis were assessed by peripheral quantitative computed tomography ( pQCT ) using the XCT 2000 device ( Stratec Medizintechnik , Pforzheim , Germany ) [62] . BLSA albumin concentrations were measured by a calorimetric assay using bromocresol green ( Ortho-Clinical Diagnostics ) . Calcium concentrations were measured by a calorimetric assay ( Vitros 5 , 1 , FS ) . CoLaus participants were genotyped using Affymetrix Human Mapping 500 K Array . For the genome-wide association stage , genotyping in LOLIPOP participants was carried out using the Illumina 317 K mapping array , Affymetrix Human Mapping 500 K array , and Perlegen , 284 K platforms ( Table S1 ) . Participants of the InCHIANTI and BLSA studies were genotyped using Illumina Infinium HumanHap 550 K SNP arrays were used for genotyping [21] . Imputation of allele dosage of SNPs was performed using either MACH [63] or IMPUTE [64] with parameters and quality control filters as described in . All European cohorts imputed SNPs typed in the HapMap CEU population; LOLIPOP Indian Asian cohort imputed SNPs using mixed HapMap populations , given that this showed greater concordance with real genotypes compared with use of any one HapMap population . SNPs were excluded if cohort-specific imputation quality as assessed by r2 . hat ( MACH ) or . info ( IMPUTE ) metrics were <0 . 30 . In total , 2 , 557 , 252 genotyped or imputed SNPs had data from one or more cohorts and were analyzed . Genotypes in deCODE were measured using either humanHap300 , humanHap300-duo or humanCNV370 .
|
Calcium levels in blood serum play an important role in many biological processes . The regulation of serum calcium is under strong genetic control . This study describes the first meta-analysis of a genome-wide association study from four cohorts totaling 12 , 865 participants of European and Indian Asian descent . Confirming previous results in some candidate gene studies , we find that common polymorphisms at the calcium-sensing receptor ( CASR ) gene locus are associated with serum calcium concentrations . We show that CASR variants give rise to the strongest signals associated with serum calcium levels in both European and Indian Asian populations , while no other locus reaches genome-wide significance . Our results show that CASR is a key player in genetic regulation of serum calcium in the adult general population .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"computational",
"biology/population",
"genetics",
"genetics",
"and",
"genomics",
"public",
"health",
"and",
"epidemiology",
"computational",
"biology/genomics"
] |
2010
|
Genome-Wide Meta-Analysis for Serum Calcium Identifies Significantly Associated SNPs near the Calcium-Sensing Receptor (CASR) Gene
|
Surveillance of antimicrobial resistance ( AMR ) is an important component of public health . Antimicrobial drug use generates selective pressure that may lead to resistance against to the administered drug , and may also select for collateral resistances to other drugs . Analysis of AMR surveillance data has focused on resistance to individual drugs but joint distributions of resistance in bacterial populations are infrequently analyzed and reported . New methods are needed to characterize and communicate joint resistance distributions . Markov networks are a class of graphical models that define connections , or edges , between pairs of variables with non-zero partial correlations and are used here to describe AMR resistance relationships . The graphical least absolute shrinkage and selection operator is used to estimate sparse Markov networks from AMR surveillance data . The method is demonstrated using a subset of Escherichia coli isolates collected by the National Antimicrobial Resistance Monitoring System between 2004 and 2012 which included AMR results for 16 drugs from 14418 isolates . Of the 119 possible unique edges , 33 unique edges were identified at least once during the study period and graphical density ranged from 16 . 2% to 24 . 8% . Two frequent dense subgraphs were noted , one containing the five β-lactam drugs and the other containing both sulfonamides , three aminoglycosides , and tetracycline . Density did not appear to change over time ( p = 0 . 71 ) . Unweighted modularity did not appear to change over time ( p = 0 . 18 ) , but a significant decreasing trend was noted in the modularity of the weighted networks ( p < 0 . 005 ) indicating relationships between drugs of different classes tended to increase in strength and frequency over time compared to relationships between drugs of the same class . The current method provides a novel method to study the joint resistance distribution , but additional work is required to unite the underlying biological and genetic characteristics of the isolates with the current results derived from phenotypic data .
The evolution of acquired antimicrobial resistance ( AMR ) in pathogenic microorganisms is one of the foremost challenges in public health today . Antimicrobial drug use in medicine and agriculture generates selective pressure that selects for AMR in bacterial populations and facilitates emergence of multiple drug resistant ( MDR ) phenotypes [1] . Bacterial pathogens with an MDR phenotype pose a substantial clinical challenge since the antimicrobial drugs typically prescribed may not effectively clear a patient’s infection and delay the patient’s recovery . The most recent and dramatic example is the emergence of plasmid-mediated resistance to colistin in Escherichia coli isolated from animals and humans [2] . Colistin is the last resort to treat some infections caused by Gram-negative bacteria such as carbapenem resistant Actinobacter baumannii and at least one pan-resistant strain of E . coli has been isolated from a clinically ill human patient [3 , 4] . Multiple mechanisms of bacterial evolution , including mutation , recombination , and clonal expansion , give rise to highly resistant strains of bacteria and encourage persistence of these strains following their emergence [5] . Genetic capitalism describes the phenomenon by which the progeny of a microbe with one fitness trait , a drug-resistant phenotype for example , tend to survive serial selection events , in turn increasing the likelihood that the progeny will acquire additional fitness traits via recombination [6] . The initial fitness trait may be acquired by mutation , horizontal gene transfer ( HGT ) , or novel recombination event with other bacteria in the environment . An example of genetic capitalism was the rapid emergence and expansion of a fluoroquinolone-resistant strain of methicillin-resistant Staphylococcus aureus ( MRSA ) at the Atlanta Veterans Affairs Medical Center , where the proportion of fluoroquinolone-resistant MRSA isolates increased from 0 to nearly 80% within 1 year of the introduction of ciprofloxacin to the hospital's formulary [7] . Collateral selection , another mechanism capable of generating MDR strains , describes the phenomenon where selection pressure from one antimicrobial drug may additionally select for or against phenotypic resistances to other drugs via several mechanisms [8] . Cross-resistance describes resistance to several related drugs by a single mechanism , e . g . , point mutations to the DNA gyrase subunit A gene ( gyrA ) that increase resistance to quinolone antibiotics [9] , co-resistance describes a set of resistances conferred by a set of frequently concurrent genes , e . g . , polymixin and colistin resistance genes carried on the recently-described MCR-1 plasmid in Escherichia coli [2] , and pleiotropic resistance describes individual mutations that simultaneously affect resistances for multiple unrelated drugs such as marR gene alterations in E . coli that increase resistance to tetracycline , chloramphenicol , and fluroquinolones drugs [10 , 11] . Monitoring the proliferation of existing MDR strains and emergence of novel MDR strains are primary goals of AMR surveillance programs carried out by governmental agencies across the world including , but not limited , to the Department of Agriculture ( USDA ) , Food and Drug Administration ( FDA ) and Centers for Disease Control and Prevention ( CDC ) in the United States [12–14] , the Department of Health in the United Kingdom [15] , the European Centers for Disease Control and Prevention ( ECDC ) in the European Union [16 , 17] , and the collaborative Transatlantic Task Force on Antimicrobial Resistance created in 2009 [18 , 19] . These agencies' survey methodologies and their published reports have been largely focused on univariate phenotypic resistances and the prevalence of MDR strains [20] . These surveillance reports however provide little information about the joint distributions of drug resistances in the overall population that may contribute to MDR strain development and emergence via collateral resistance . Outside of AMR surveillance , several studies of in vitro and in situ bacterial populations have begun to explore joint resistance distributions in an effort to identify strategies to mitigate MDR development [21–23] . Studying joint distributions of resistances in AMR surveillance data is necessary to understand in situ MDR strain evolution , but poses a number of challenges . The number of estimated correlation coefficients required to fully describe the pair-wise resistance distributions grows quadratically with the number of drugs in a panel , specifically kC2 = k2/2 –k/2 for a panel of k drugs; Most AMR panels contain a dozen drugs or more , requiring at least 12C2 = 66 correlations to be estimated . Hypothesis testing may be used to determine which drug pairs are not correlated , but the large number of estimated correlations results substantial type I error rate inflation . Pair-wise correlations also do not control for confounding by other variables in the dataset . Better methods are needed to characterize and quantify the joint resistance distributions in bacteria . We propose graphical models , specifically Markov networks , to study joint distributions in AMR surveillance . Graphical models are mathematical constructs which represent the interactions between elements in complex systems . A variety of well-studied parameters , e . g . , density and modularity , have been described to summarize these models at multiple levels of complexity . This study’s objective is to estimate the Markov networks' structures that represent existing AMR surveillance data to effectively describe and visualize AMR relationships among a population of bacteria . This technique is intended to supplement the current methods used to analyze and report AMR surveillance results . The method is demonstrated here using data collected by the National Antimicrobial Resistance Monitoring System for Enteric Bacteria ( NARMS ) .
The NARMS study tracks AMR in E . coli , Salmonella spp . , Enterococcus spp . , and Campylobacter spp . isolated from slaughter houses , retail meat , and cases of food borne illness by the USDA , FDA and CDC respectively . Surveillance results from NARMS between 1998 and 2013 were recently made publically available [24] . Isolates were tested for AMR with genus-specific drug panels , typically containing 13 to 15 drugs with resistance results reported as minimum inhibitory concentrations ( MIC ) . To demonstrate our approach , a subset of the NARMS data was selected from the AMR results from E . coli isolated between 2004 and 2012 ( n = 14 , 418 ) . The MIC results for the following 16 of the available 23 drugs were used to demonstrate the networks: ampicillin ( AMP ) , amoxicillin and clavulanic acid ( AMC ) , ceftriaxone ( AXO ) , cefoxitime ( FOX ) , ceftiofur ( TIO ) , amikacin ( AMI ) , gentamicin ( GEN ) , kanamycin ( KAN ) , streptomycin ( STR ) , nalidixic acid ( NAL ) , ciprofloxacin ( CIP ) , sulfisoxazole ( FIS ) , trimethoprim and sulfamethoxazole ( COT ) , chloramphenicol ( CHL ) , tetracycline ( TET ) , and azithromycin ( AZI ) . The drug resistances were grouped into classes based on the structure of the drug tested ( Table 1 ) . The breakpoints for these antimicrobials and summary of observed resistances in the NARMS data are provided in Table 2 . Additional information regarding the data and sample sizes is available in supplemental material ( S1 Text ) . All statistical analyses were performed using R version 3 . 2 . 3 [25] . Spearman's rank correlations were used to estimate MIC relationships and trends in graphical parameters over time [26] . Sparse graphical model structures were constructed using the glasso package version 1 . 8 [27] and graphical parameters were estimated using the igraph package version 1 . 0 . 1 [28] . A graphical model G of a system is comprised two sets: the vertex set G ( V ) which defines the system’s k discrete elements and the edge set G ( E ) which defines the m pair-wise interactions between the system’s elements . For simple graphical models , G ( E ) may be expressed as the set of unique , unordered pairs of adjacent vertices ( G ( E ) = { ( vi , vj ) | vi adjacent to vj , i ≠ j} or as an adjacency matrix [29] . A Markov network is a specific type of undirected graphical model used to represent relationships between variables in a data set [30 , 31] . Each of the dataset’s k constituent random variables are represented by a vertex in the Markov network ( G ( V ) = {vi , … , vk} ) . Edges in a Markov network are defined by the variables' partial correlations ( ωij ) , which represent the standardized covariance of a pair of variables vi and vj when conditioned on all other variables in the dataset . When ωij = 0 , vi and vj are conditionally independent and are not adjacent in a Markov network . Therefore , a Markov network's edge list consists of the m unique , unordered pairs of variables which are not conditionally independent ( G ( E ) = { ( vi , vj ) | ωij ≠ 0} ) and the network’s adjacency matrix A may be defined from Ω using the indicator function 𝟙M ( Eq 1 ) . When the structure of a system's Markov network is unknown , it is possible to estimate the partial correlations and network structure from observed data . An empirical correlation matrix Σ can be inverted to produce an precision matrix Θ ( Θ = Σ-1 ) , which in turn can be used to estimate Ω ( Eq 2 ) [32] . Graphical models estimated in this way are typically complete ( m = kC2 ) since trivial correlations will persist , even after conditioning on the other variables in the set . However , graphical models , including Markov networks , are easier to interpret and are more useful when they are sparse ( m << kC2 ) [31] . The least absolute shrinkage and selection operator ( LASSO ) is a method used to generate more parsimonious models of joint distributions . The graphical LASSO applies an L1-regularization penalty ρ to estimate a penalized precision matrix ( Θ* ) [27] . The penalization reduces the absolute value of all elements of Θ ( |ϑij| > |ϑ*ij| ) and if the penalty is large enough , some ϑij* are reduced to zero . When estimated from Θ* instead of Θ , the trivially small partial correlations are reduced to zero and the Markov network will be sparse . While any non-negative value may be assigned to ρ , its useful range is limited to min|ϑij|< ρ < max ( |ϑij| ) . When ρ > max ( |ϑij| ) , Θ* and Ω are diagonal , all ωij are trivial , and G ( E ) = ∅; conversely , if ρ < min ( |ϑij| ) , then no ωij are trivial , and the resultant graphical model will typically be complete , similar to using the unpenalized precision matrix . The Markov networks of collateral resistance developed here are subsequently referred to as "R-nets" and two versions are presented: simple R-nets ( R ) and weighted R-nets ( R’ ) . For these networks , R ( V ) = R’ ( V ) and R ( E ) = R’ ( E ) , but their adjacency matrices are different . For R , the simple adjacency matrix A is defined by Eq 1 . For R’ , Ω is used as the weighted adjacency matrix where the elements ωij represent the strength and direction of the relationship . The vertices of an R-net represent the observed distribution antimicrobial resistances , e . g . , TIO represents the observed set of MICTIO . The edges of an R-net represent correlated resistances , which in turn identify potential collateral resistances in the population . A Spearman's rank correlation matrix Σy was estimated for the adjusted log2MIC values within each year y . A non-parametric estimator was used because the distribution of the MICs , even when transformed , frequently did not conform to a normal distribution , and Pearson’s correlation estimates may yield biased results when applied to non-normal data [26] . The graphical LASSO and Eq 2 were used to estimate the structures of Ry and R’y . The graphical LASSO method assumes multivariate normal data , but it has been demonstrated that an untransformed Spearman’s correlation matrix may be substituted for the more traditional parametric correlation matrix , provided the former is positive semidefinite [33] . The structures of sparse Markov networks generated via the graphical LASSO are conditional on the L1-penalty used to estimate Θ* , thus the selection of the penalty is an important step in the inference process . In general , the value of ρ should be low enough to assure no important edges are lost , and high enough to effectively reduce the density of the graph . Comparisons of Ry and R’y over time were restricted to sets of R-nets generated from a common penalty to improve comparability of the networks . To select a common penalty value , 12 sets of R and R’ were produced and evaluated with each set of networks generated using a separate value of ρ between 0 . 05 and 0 . 60 . The structure and distribution of density ( m¯y; Eq 3 ) within each set of R-nets were subjectively evaluated to select a single penalty value to apply across all years . The common penalty was selected via a supervised review of the R-nets based on network interpretability and trends in m¯y over ρ . A network should be sparse enough to be understood when visualized , meaning the graph should not be too dense for a viewer to understand the set of edges present . Edges representing expected cross-resistances , e . g . , NAL-CIP , AMP-AMC , and FIS-COT , should be strong and frequently present . Additionally , the number of unstable cycles ( cycles with an odd number of negative edges ) should be minimized [34] . The median density ( m¯50 ) and range of m¯y was plotted against the evaluated values of ρ and a point was sought where the slope of the line broke from steeply decreasing to level off; this method is similar to the Scree test for factor retention in FA and PCA [35 , 36] . The lowest penalty that generated R and R’ that met these criteria were selected for further evaluation . Three graphical parameters were used to evaluate changes in AMR resistance over time described by R and R': density , vertex degree , and modularity . Density quantifies a graph's overall interconnectivity and trends in m¯ could indicate overall trends in AMR relationships and risk of collateral resistance in the population over time [37] . Increasing density over time represents more interconnectivity of drug resistances . Vertex degree di is equal to the number of other vertices adjacent to vi . In an R-net , d describes the number of other resistances one MIC is related to in the population . In the R-nets vertices with high di may represent resistances which are influential to the development of MDR strains . For example , if resistance to drug A had a high degree in a bacterial population , e . g . , dA = 5 , the use of drug A could select for increased resistance to A and potentially affect resistance to five other drugs , assuming that the covariance structure of phenotypic drug resistance identified by the Markov networks reflects genetic mechanisms allowing for collateral resistance . The high vertex degree of A does not indicate that the drug resistance is responsible for resistance noted in five adjacent vertices , only that selection for A could possibly influence the other resistances . Modularity ( Q ) measures how frequently adjacent vertices are similar or dissimilar as defined by vertex attributes [38] . The class of drug associated with the resistance of each vertex was used to assign group membership ( Table 1 ) . Modularity is positive when edges join similar vertices more frequently than would be expected by chance and negative when dissimilar vertices are more frequently than would be expected by chance . Trends in modularity may represent shifting tendencies in resistance relationships to exist between more similar or less similar drugs . Similar estimates of modularity for signed and weighted networks ( Q' ) can be estimated [39] . The value of Q' describes the relative strength of edges between similar vertices compared to dissimilar vertices . Spearman's rank correlation was used to test for a trends in m¯ , Q , and Q' over time due to the small number of values being compared [26] . Principal component analysis has been used to study resistance relationships and was performed here to evaluate how the current method compares to a previously employed method [23] . Three years of the study were selected to represent data from the beginning , middle , and end of the study , respectively , and PCA was performed separately on the data from each year . The eigenvalues of Σy ( λ ) were computed and the components for which λ > 1 were extracted and subsequently oblimin rotated [40 , 41] . The oblimin rotation , an oblique rotation method , was used to allow the loadings to assort naturally without imposing orthagonality . Drug resistances with component loadings that had an absolute value greater than 0 . 4 were considered to be important and were assigned to the respective rotated component [42] . The rotated components were compared to the respective R-nets to evaluate agreement between the two methods .
The range of ρ evaluated was fully contained in the useful range identified above , with 0 < |ϑij| < 2 . 7 for all years in the study . The upper region of the possible penalties ( 0 . 6 < ρ < 2 . 7 ) was not evaluated since the networks generated in this region would have produced networks too sparse to be informative . The density of R-nets over all ρ values and years ranged from 4 . 8% to 57 . 1% over the evaluated range of ρ , and 0 . 3 ≤ ρ ≤ 0 . 5 generated graphs with very similar densities between 10% and 20% , indicating consistent graph structures in this range of ρ ( Fig 1 ) and changes in the slope of m¯50 over ρ were noted at ρ = 0 . 10 and ρ = 0 . 25 ( Fig 2 ) . In general , the R-nets generated by ρ = 0 . 10 were too dense to easily interpret and several unstable cycles were noted in R’2008 , R’2009 and R’2010 . The R-nets generated by ρ = 0 . 25 were sparse enough to be interpreted with reasonable effort . Additionally , R’ under ρ = 0 . 25 contained no unstable cycles since all partial correlations were positive . The latter penalty of ρ = 0 . 25 was selected as the common penalty and used to generate R and R’ interpreted below ( Fig 3 ) . The supplemental material provides an overview of ρ’s impact on density and modularity ( S2 Text ) and a more in-depth description of R and R’ conditioned on ρ = 0 . 10 ( S3 Text ) . A total of 119 unique edges could have observed during the entire study period ( the AZI-AMI edge could not have been observed since both drugs were never included in the same year ) , and 105 edges could have been observed in each year . Of these 119 unique edges , 33 unique edges were observed in at least one year of the study , and 15 appeared in all 9 years . A bimodal distribution was noted in the frequency of edge appearance , with one group of edges observed in 1 to 4 years , and another group of edges observed for 6 years or more ( Fig 4 ) . Sixteen of the 33 observed edges found represented resistance relationships between drugs of the same class ( blue areas of Fig 4 ) , most of which were present in all nine years during the study ( 13/16 within-class edges ) indicating the relative stability of these relationships during the study period ( Fig 5A ) . Seventeen between-class edges ( red areas of Fig 4 ) were identified , of which only 4 were present for 8 or 9 years . The majority of between-class edges ( 12/17 between-class edges ) were present for 3 years or less ( Fig 5B ) . No significant trend in m¯ over time was noted ( Spearman’s rho = -0 . 14 , p = 0 . 71 ) . Two dense subgraphs of R were noted to be frequently present . The first frequent dense subgraph RBLA contained the five β-lactam drug resistances AMC , AMP , AXO , FOX , and TIO and formed a clique in every Ry . The second frequent dense subgraph RAST included GEN , KAN , STR , COT , FIS and TET , and m¯AST exceeded 50% in six out of the nine years . Individual vertices demonstrated several different patterns of degrees during the study period , and d ≤ 6 for all vertices and time periods . For the β-lactam drugs , d = 4 was typical but occasionally increased to 5 or 6 . Different patterns were noted among the degrees of the aminoglycosides: dKAN was high the earlier in the study and decreased in later years ( 3 ≥ dKAN in 2009 and prior and 2 ≤ dKAN in 2010 and later ) , dGEN and dAMI were more consistent , and dSTR varied widely and without a clear pattern . In every time period , dCIP = 1 , corresponding to its edge with NAL , but dNAL tended to increase through the study . The values for dCHL , dFIS , dCOT , and dAZI changed little during the study period . There was substantial variation in dTET and did not appear to follow a pattern . Modularity defined by class in the unweighted R-nets ranged from Q2012 = 0 . 267 to Q2005 = Q2006 = 0 . 420 . Modularity was not significantly associated with time ( Spearman's rho = -0 . 49 , p = 0 . 18; Fig 6 ) . When weighted by partial correlation , modularity estimates ranged from Q'2012 = 0 . 321 to Q'2005 = 0 . 466 , and a statistically-significant decreasing trend was noted in Q' ( Spearman's rho = -0 . 95 , p < 0 . 005 ) . Principal component analysis was performed on Σ2004 , Σ2008 , Σ2012 . In each of the three years , four components were identified for extraction and oblimin rotation . Combined , these rotated components accounted for 61% , 64% , and 61% of the overall variance in log2MIC in 2004 , 2008 , and 2012 respectively . The loadings for each component were similar , though not identical , over all three years ( Table 3 ) . Correlations between the rotated components were small , with -0 . 10 ≤ r ≤ 0 . 25 . Most rotated components aligned with subgraphs with m¯ > 50% ( Fig 7 ) .
The representation of AMR surveillance data using Markov networks generated via the graphical LASSO is a novel method to characterize potential collateral resistances in bacterial populations . The graphical nature of this method lends itself to simple visualization which allows complex relationships to be communicated clearly . The structures of R2004 , R2008 , and R2012 are similar to the variance structures identified by the respective PCA results , but the R-nets provide results that are simultaneously more detailed and more interpretable than the results from PCA . Several resistance relationship patterns appeared in the models over time . Resistances to the β-lactam drugs were consistently and strongly related to each other in all years , generating the complete induced subgraph RBLA . These patterns of related classes are likely an example of cross-resistance , though pleiotrophic and co-resistance mechanisms may also be present . The elements of RAST ( V ) did not represent resistances to a single class of drugs or even drugs targeting a common metabolic pathway . A similar grouping of resistances without GEN was previously described in beef cattle , where it was attributed to antibiotics frequently used in production medicine [23] . The quinolone drugs were always correlated with each other , and while CIP was only correlated with NAL , NAL was additionally correlated with CHL , TIO , and AMP in some years . Additionally , these patterns are consistent with those seen in the PCA of the same data , and similar to patterns seen in a previous study of AMR relationships in E . coli [23] . Evaluation of the graphical parameters provided additional insight into changes in the joint distributions of resistance over time . No temporal trend was apparent in m¯ indicating that , on average , the amount of interconnectivity of AMR in this population of E . coli did not change substantially over time . The negative trend noted in Q' indicates a shift towards stronger relationships between drugs of different classes , weaker relationships between drugs of the same class , or both . Visual evaluation of Fig 5A and 5B suggests both processes may be occurring , with within-class edges becoming slightly weaker over time and the appearance of more , stronger between-class edges . Lower weighted modularity values over time were consistent across the common penalties and may indicate a concerning increase in co-resistance and pleiomorphic mutations in this population of E . coli . Neither the source of the decreasing trend in Q’ or the behavior of any specific edge can be assigned without isolate covariate data , which were not available for data from the NARMS study . Quantification of specific network structures into simple numeric criteria is one of the major advantages of graphical models . While some information about a graph’s structure is lost when the structure is condensed into a simple criterion , these criteria greatly facilitate the comparison of graphs . Parallel interpretation of multiple criteria , as is done here with Q , Q’ , and m¯ , can provide a more complete description . It should be emphasized that the R-nets describe the joint relationships of resistances at the population level , hence little can be inferred about the genotype or the phenotypic resistance profile of an individual isolate from these results . At best , a probabilistic statement can be made about some MIC values given knowledge about the others . It is also not currently possible to infer the source of the resistance or resistance relationships based on these models . The edges of the R-nets represent potential resistance relationships , but existence of the edge alone is not sufficient to induce collateral resistance . For example , even though GEN and FIS were adjacent , GEN will not affect FIS unless there exists a concurrent selection pressure for or against gentamicin resistance . Induction of collateral resistance for FIS by GEN requires the combination of the FIS-GEN resistance relationship and the selection for GEN , potentially from the therapeutic use of gentamicin . This illustrates that knowledge of the resistance relationships alone is insufficient to determine how the R-nets have influenced the patterns of resistance observed; antimicrobial use data are also needed and combining R-net results with exposure data is a topic of ongoing research . Despite this limitation , the R-nets can still identify phenotypic resistances of interest prospectively , specifically vertices of relatively high degree . Application of the methods to multiple strains from varied sources will allow the opportunity to objectively identify resistance patterns , and potentially enable predictions about resistance relationship evolution based on previously observed R-nets . There are several applications where the population-level focus of the currently described method would be useful . First , this method could be directly applied to AMR surveillance in health care facilities where local evolution of highly resistant bacterial strains is a major concern [7] . While the dataset used to demonstrate the current method was relatively large and n >> m even in the unpenalized case , the graphical LASSO can perform well even when the number of parameters to estimate exceeds the number of available observations [43] . Therefore , the current method may be employed to estimate R even when the sample size is modest , as may be the case in a single health care facility . R-nets can be estimated retrospectively from clinical data to monitor for antimicrobials with a high degree , indicating a high risk of extensive collateral resistance . If drug use data were available from patient records , limited inference could be made about how an MDR strain evolved in a clinical environment . If appropriate negative correlations were noted , a selection inversion could also be attempted [21] . The R-nets could also be applied to monitor emergence of novel resistance relationships at any scale over time . Knowledge of resistance relationship dynamics will help improve clinical decision making by informing the physician about what resistances may altered by use of a specific drug . The R-nets can help inform policy making by similarly tracking resistance relationships and detecting resistance of relationships of concern , and can help facilitate AMR research by screening large numbers correlations to locate non-trivial associations . The R-nets revealed several interesting patterns of AMR involving the resistances the quinolones drugs NAL and CIP . In E . coli , quinolone resistance may be increased by mutations in the A and B subunits of DNA gyrase [9 , 44] and mutations in parC and parE , which encode subunits of DNA topoisomerase IV[45–47] . Increases in NAL and CIP due to these mutations would be expected to lead to a strong relationship between NAL and CIP , but no other drugs in the panel since the quinolones are the only class of antimicrobials affected by DNA gyrase and topoisomerase mutations . This relationship was observed , with ωNAL , CIP > 0 . 1 in most years indicating a consistent strong NAL-CIP edge . Increased expression of active , non-specific efflux pumps in E . coli is an alternate and complementary mechanism of quinolone resistance , and also increases resistance to chloramphenicol , tetracycline , and other classes of drugs [10 , 11 , 48] . If efflux pumps were an important mechanism of quinolone resistance in the sampled population of E . coli , correlations should be noted among NAL , CIP , TET and CHL and a dense subgraph including these vertices should be noted . However , no such subgraph was noted: TET was never found to be correlated with CIP , NAL , or CHL , and ωCHL , NAL < 0 . 05 most years . Efflux pump expression may explain the weak CHL-NAL edge but , overall , these results would appear to support the conclusion that efflux pumps expression was not an important source quinolone resistance , or at least was not a major contributor to resistance in this population . Given ρ = 0 . 25 , every edge in every year represented a positive partial correlation in resistances , indicating that , on average , increasing resistances to one drug was only associated with increased , and never decreased , resistances of the adjacent vertices . Among the unpenalized partial correlations , positive partial correlations outnumbered negative partial correlations by about a 2-to-1 margin , and the median absolute value of negative partial correlations was about half that of positive correlations . Negative partial correlations were only noted when ρ ≤ 0 . 15 . These findings are consistent with the phenomenon of genetic capitalism where the progeny of bacteria with at least one advantageous mutation tend to acquire other additional advantageous traits over time via recombination and HGT [6] . The relative weakness and infrequency of negative partial correlations compared to positive partial correlations is also consistent with the patterns seen in the resistance relationship networks of previous studies [22 , 49] . One application of R-nets and similar networks is to identify pairs or larger groups of collaterally susceptible antibiotics to create a selection inversion: a reduction in overall AMR via strategic antibiotic use [21 , 22] . Without any negative partial correlations , it is unlikely that a selection inversion could be achieved in this population of E . coli , but could be feasible in other populations . The purpose of the L1 penalization in the graphical LASSO procedure is to eliminate trivial edges from the graph by reducing their corresponding elements of Ω to zero , but the penalization also biases the non-trivial elements towards zero , as well [27 , 50] . This is effect is the ‘shrinkage’ aspect of the least absolute shrinkage and selection operator . For example , the estimated unpenalized partial correlation between AMC and AMP in 2012 is 0 . 62 , but when estimated via the graphical LASSO with ρ = 0 . 25 this estimated partial correlation is reduced to 0 . 36; the latter estimate of ωAMC , AMP has been moved towards 0 , or shrunken , compared to the former due to the L1 penalization . This bias caused by the penalization should be kept in mind when interpreting the magnitudes of the edge weights since edges attributed to small penalized partial correlations may actually correspond to substantially larger partial correlations in the unpenalized matrix . The penalized partial correlations are still useful , though , because they provide a useful method to compare the relative strengths edges within a graph or graph series . Though R ( E ) = R' ( E ) , the unweighted models are not affected by this bias since A does not incorporate magnitude of the partial correlations . Hence , R may be preferable to R' for some applications . One important limitation of Markov networks is that they cannot specifically identify the higher order dependence structure of joint distributions complete induced subgraphs [31] . Without additional information about the population or the random variables comprising by R ( V ) , the researcher must rely on induction to determine which covariance structure is more likely to be correct . For example , a complete induced subgraph of a Markov network with k = 4 could be the result of six separate pair-wise correlations , a combination of one 3-way and 3 pair-wise correlations , or a single 4-way correlation . The frequent dense subgraph RBLA is a case in the current study where a higher order dependence structure may exist but cannot be specified . Due to the strength of the edges , consistency of edges over time , and similarity of the drugs in RBLA , it would be reasonable to attribute the observed structure to a single 5-way correlation element , but combinations of multiple lower order interactions are possible . The R-nets are unable to distinguish the biochemical or genetic mechanisms responsible for the observed phenotypic relationships , and individual edges may represent multiple categories of collateral resistance . However , some inferences may be reached via induction based our biological knowledge of the resistance mechanisms . Edges between drug resistances in the same class could be caused by cross-resistance mechanisms , or pleiotropic mutations affecting the common site of multiple drugs . For example , it is likely that β-lactamase enzymes , variations in penicillin-binding proteins or both are responsible for the dense subgraph of AMP , AMC , AXO , FOX , and TIO[51 , 52] . Edges between drug resistances that belong to different classes could be caused by either co-resistance or pleiotropic resistance separately . Multiple different mechanisms may be present active in a single R-net , and multiple types of collateral resistance may contribute to a particular edge . The AZI-CHL edge was observed both years where AZI ∈ Ry ( V ) , and both drugs target the 50S ribosomal subunit . This relationship could be attributed to a pleiotropic mutation of the 50S subunit affecting the action of both AZI and CHL , co-resistance from genes residing on a plasmid or in the same genetic cassette , or both . Genotypic data could be used to elucidate the mechanisms underlying the observed phenotypic correlations , and the structure of R may provide insight into the genetic mechanisms leading the MDR bacteria . An important step in the validation of the current method will be to demonstrate the structure of R generated by a bacterial population conforms to the combinations of resistance genes present in the sampled bacteria . The selection of ρ in the current method is of particular importance due to its influence on the generated networks . Here , ρ = 0 . 25 was selected because this value produced biologically coherent graphs in an informative range of densities around 20% . Larger or smaller values of ρ , such as the scenario presented in S3 , may be appropriate for other applications . The subjectivity of the selection is mitigated since Θ is uniformly penalized , so any edges that are trivial at a given value of ρ will also be trivial at all higher values . Future work will explore additional methods for selecting ρ to improve standardization across studies . The graphical nature of this method lends itself to simple visualization , allowing for complex relationships to be communicated clearly and additionally provides a framework for further analysis where the presence and magnitude of the partial correlations provide an outcome for evaluation via other statistical methods . Continuous covariates , including the number of gene copies present in an isolate , could be included as nodes in an R-net in addition to the drug resistances . Edges between a covariate node and resistance nodes could identify relationships similar to collateral resistance . Covariates of high degree may identify avenues for the indirect induction of AMR similar to collateral resistance , albeit indirectly . Graphical parameters for these covariate nodes such as centrality could provide additional insight into the evolution of highly resistant strains . A separate strategy for comparing R-nets is to test for differences in Ω generated by bacterial subpopulations . A number of methods for comparison of covariance matrices have been proposed [53] and may present a feasible proxy method to compare R’ structure since R’ is defined by Ω . Samples can be stratified using isolate date , e . g . , isolate source , time interval of collection , genotype , etc . , R’ estimated for each stratum , and Ω compared across the strata . Unlike the strategies for comparison of covariates described earlier , which included covariates as nodes within the network , Ω comparisons estimate separate R-nets based on covariates external to the network and therefore must identify covariates of interest a priori . Additional work is needed to determine how to best evaluate covariates in R-nets , and it will be important to validate the phenotypic findings against genetic data to improve the interpretability of results . We chose evaluate joint distributions of log-transformed MIC values , but it is common practice to dichotomize MIC values into susceptible or resistant categories based on breakpoints when analyzing resistance data [14 , 49 , 54] . The transformed AMR data contain more information than dichotomized results , and therefore analyses of the continuous data is more powerful than similar analyses of dichotomized data [55 , 56] . The dichotomization of MIC values is also dependent on the selected breakpoints , which are based on clinically-relevant drug concentrations , and not based on the distribution of MICs within bacterial populations [57] . Breakpoints may also vary over time and geographic region [57–61] . We believe the networks developed from continuous resistance data more accurately represent the resistance relationships than networks based on more traditional dichotomized results . Previous studies have used directed graphical models to describe the structure joint distributions of resistance in E . coli [22 , 49 , 62] . Directional edges represent types of causal relationships in the system they represent . In some cases , it is appropriate to assign directionality to AMR relationships , e . g . , collateral resistance generated by in vitro selection for resistance to individual drugs [22] or the association between genes and phenotypical resistances [62] . In many other situations , including observational or surveillance studies of AMR where antimicrobial use is unknown , causality of an AMR relationship , if any exists , must be assumed [49] . In contrast , Markov networks are undirected and have no implied causality , which avoids the risk of incorrect assumptions regarding cause and effect . While effective for describing and visualizing AMR relationships , the R-nets cannot provide information about univariate changes to the MICs over time . Hence , this method is intended complement current surveillance methodologies , not replace them . It was noted that the majority of MICs test ranges included at most 1 dilution above the CLSI suggested breakpoint for resistance , and some MIC test ranges had as few as two dilutions . Increasing the number of dilutions tested for MICs would capture more information and increase the accuracy of the R-nets .
The graphical models presented provide a novel method of mapping resistance relationships in observed in AMR surveillance data . The R-nets present a powerful and useful tool can provide insight into the evolution of MDR bacterial strains and allow simple visualization of complex AMR data . Future work is needed to validate the R-nets against genotypic results to confirm the observed phenotypic resistances accurately represent the underlying biochemical mechanisms of AMR . The application of the method to pathogenic species of bacteria , including MRSA , Salmonella spp . and Campylobacter spp . , is planned and may provide insight into antimicrobials driving the evolution and emergence MDR bacteria .
|
Surveillance of antimicrobial resistance patterns is an important responsibility in modern public health . Due to the genetic configuration of bacterial pathogens , the use of a single antimicrobial drug may select for bacteria that are resistant to multiple other antimicrobial drugs via set of mechanisms collectively known as “collateral resistance” . We have developed a new analytic method to study and describe possible collateral resistance pathways using existing antimicrobial resistance surveillance data . The method , named “R-nets” , use network models to visualize the patterns of collateral resistance which allows large amounts of information to be clearly communicated to public health officials and researchers . Applying the R-nets to publicly available data from E . coli collected by the FDA and USDA between 2004 and 20012 we found that the number of collateral resistance links was relatively constant , but there may be an increase in collateral resistances between drugs of different structures . The method described will hopefully be useful for predicting and managing the proliferation of highly resistant bacterial pathogens
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion",
"Conclusions"
] |
[
"bacteriology",
"antimicrobials",
"livestock",
"organismal",
"evolution",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"drugs",
"microbiology",
"vertebrates",
"animals",
"antibiotic",
"resistance",
"microbial",
"evolution",
"pharmacology",
"bacterial",
"pathogens",
"antimicrobial",
"resistance",
"birds",
"medical",
"microbiology",
"microbial",
"pathogens",
"gamefowl",
"drug",
"distribution",
"fowl",
"pharmacokinetics",
"agriculture",
"poultry",
"bacterial",
"evolution",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"chickens",
"evolutionary",
"biology",
"drug",
"interactions",
"amniotes",
"organisms"
] |
2016
|
Markov Networks of Collateral Resistance: National Antimicrobial Resistance Monitoring System Surveillance Results from Escherichia coli Isolates, 2004-2012
|
The factors that govern the development of tuberculosis disease are incompletely understood . We hypothesized that some strains of Mycobacterium tuberculosis ( M . tuberculosis ) are more capable of causing disseminated disease than others and may be associated with polymorphisms in host genes responsible for the innate immune response to infection . We compared the host and bacterial genotype in 187 Vietnamese adults with tuberculous meningitis ( TBM ) and 237 Vietnamese adults with uncomplicated pulmonary tuberculosis . The host genotype of tuberculosis cases was also compared with the genotype of 392 cord blood controls from the same population . Isolates of M . tuberculosis were genotyped by large sequence polymorphisms . The hosts were defined by polymorphisms in genes encoding Toll-interleukin 1 receptor domain containing adaptor protein ( TIRAP ) and Toll-like receptor-2 ( TLR-2 ) . We found a significant protective association between the Euro-American lineage of M . tuberculosis and pulmonary rather than meningeal tuberculosis ( Odds ratio ( OR ) for causing TBM 0 . 395 , 95% confidence intervals ( C . I . ) 0 . 193–0 . 806 , P = 0 . 009 ) , suggesting these strains are less capable of extra-pulmonary dissemination than others in the study population . We also found that individuals with the C allele of TLR-2 T597C allele were more likely to have tuberculosis caused by the East-Asian/Beijing genotype ( OR = 1 . 57 [95% C . I . 1 . 15–2 . 15] ) than other individuals . The study provides evidence that M . tuberculosis genotype influences clinical disease phenotype and demonstrates , for the first time , a significant interaction between host and bacterial genotypes and the development of tuberculosis .
It is estimated that one third of the world's population is infected with Mycobacterium tuberculosis ( M . tuberculosis ) , although the majority will never develop active disease . The factors that govern the development of tuberculosis disease are complex and incompletely understood . Various factors have been clearly associated with increased susceptibility to tuberculosis . HIV infection is by far the most important; it increases the lifetime risk of sub-clinical infection converting to active disease from 1 in 10 to 1 in 3 [1] and is strongly associated with disseminated disease . Defining the contribution of host genetic polymorphisms to disease susceptibility has been more difficult . Studies have suggested polymorphisms in several genes are associated with the development of pulmonary tuberculosis . Some of the genes with polymorphisms that have been validated in multiple studies and may have an effect on gene function include solute carrier family 11 , member 1 ( SLC11A1 , formerly NRAMP1 ) [2]–[6] , interferon gamma [7] , [8] , TIRAP/MAL [9] , P2XA7 [10] , [11] , and CCL2 ( or MCP-1 ) , [12]–[14] . Others have shown the less common extra-pulmonary manifestations of tuberculosis may have a different host genetic susceptibility profile and have implicated various polymorphism in components of the innate host response to infection [15] [16] , [17] [18] , [19] . We have recently reported associations between the development of TBM and single nucleotide polymorphisms ( SNP ) in the Toll-interleukin-1 receptor domain containing adaptor protein ( TIRAP ) and Toll-like receptor-2 ( TLR-2 ) genes [19] , [20] . However , tuberculosis disease results from the interactions between host and bacteria and there have been no studies examining the influence and relationship of both host and bacterial genotype variation on clinical disease phenotype . M . tuberculosis exhibits a clonal population structure [21] , [22] and therefore was regarded until recently as an organism with little relevant genetic variation [23] . However , studies examining M . tuberculosis isolates from wider geographic distributions using whole genome scanning approaches have revealed a cladal phylogeographic distribution with significant variation between major lineages , each of which is associated with specific geographic regions [24] , [25] ( Figure 1 ) . The degree to which this genetic variation influences disease phenotype has been difficult to study . In vitro and in vivo models of infection have shown different genotypes of M . tuberculosis induce different patterns of host immune response [26]–[30] , but the relevance of these findings to human disease remains uncertain . Epidemiological studies have found some genotypes may be associated with different disease phenotypes . For example , several studies have suggested an association between mycobacterial plc gene polymorphism and disseminated extra-pulmonary disease [31]–[33] , but these studies have been small , retrospective , or unable to determine if differences are due to host genetic susceptibility or bacterial genetic virulence determinants . There has been much interest in the Beijing genotype of M . tuberculosis , which is highly prevalent in Asia and the states of the former USSR and has been responsible for outbreaks of multi-drug resistant tuberculosis in the USA [23] , [34] . Animal models of infection with this genotype have suggested it leads to a hypervirulent phenotype compared with other common strains of M . tuberculosis [35] . This behaviour has been attributed to an intact polyketide synthase ( pks 15/1 ) gene and the production of a phenolic glycolipid ( PGL ) [29] . PGL synthesis appears to attenuate the early host immune response to infection and is associated with reduced production of inflammatory cytokines ( 30 ) . The ability of Beijing strains to elude the host innate immune response may explain why a recent study has found this genotype is associated with haematogenously disseminated disease [36] . Animal infection models suggest haematogenous dissemination of infection occurs before the onset of T-cell mediated immunity [37] and supports the hypothesis that the ability of different strains of M . tuberculosis to produce different clinical phenotypes varies dependent upon their interaction with the host innate immune response . The study described here examined the relationship between polymorphisms in genes responsible for host innate immunity , bacterial genotype , and the development of pulmonary or meningeal tuberculosis . TBM represents the most severe form of haematogenously disseminated tuberculosis causing death or severe disability in more than half of sufferers [38] . We demonstrate that bacterial genotype does influence disease phenotype and interactions between bacterial and host genotype further influence disease expression .
The polymorphisms found in the TIRAP and TLR-2 genes and their associations with disease phenotype have been reported previously [19] , [20] . In brief , we found previously that the TIRAP SNP C558T and the TLR-2 SNP T597C were associated with susceptibility to meningeal rather than pulmonary tuberculosis and this was reconfirmed in the current dataset . Therefore , we examined whether these polymorphisms were associated with infection with any particular bacterial genotype and whether the relationship influenced disease phenotype . Host genotype was available on 314 patients; TIRAP 558 genotype was defined in 313 ( 145 TBM , 168 pulmonary ) and TLR2 597 in 306 ( 141 TBM , 165 pulmonary ) . The polymorphism frequencies and pathogen genotypes are shown in Table 4 . All SNPs were in Hardy Weinberg equilibrium ( HWE ) in cord-blood control individuals ( P≥0 . 05 ) . We analyzed the distribution of alleles and genotypes of the TB groups in comparison with the cord-blood controls ( Table 4 ) . TIRAP C558T was associated with susceptibility to TBM as previously reported OR = 2 . 96 [95% C . I . 1 . 71–5 . 11] , however , there was no stronger association between TIRAP C558T and TB caused by any unique M . tuberculosis lineage ( data not shown ) . As previously reported [20] , the TLR2 T597C polymorphism was associated with all cases of tuberculosis ( control vs . all isolates; OR = 1 . 28 [95% C . I . 1 . 01–1 . 62] , P = 0 . 045 ) . However , the allelic association was strongest for TB cases caused by the Beijing genotype isolates ( control vs . East Asian/Beijing; OR = 1 . 57 [95% C . I . 1 . 45–2 . 15] , P = 0 . 004 ) . There was no association between the TLR2 597C polymorphism and tuberculosis caused by the Indo-Oceanic ( P = 0 . 457 ) and Euro-American isolates ( P = 0 . 505 ) . We next examined whether clinical disease phenotype , pulmonary or meningeal disease , influenced the association between TLR2 T597C and bacterial genotype . There was no allelic association between TLR2 T597C and pulmonary TB caused by non-Beijing isolates ( control vs . pulmonary non-Beijing: OR = 1 . 00 [95% CI 0 . 71–1 . 43] P = 0 . 991 ) or for Beijing isolates ( control vs . pulmonary East-Asian/Beijing: OR = 1 . 27 [95% C . I . 0 . 82–1 . 47] , P = 0 . 264 ) ( Table 4 ) . There was an overall association of TLR2 T597C with meningeal disease ( OR = 1 . 51 [95% C . I . 1 . 12–2 . 03] P = 0 . 006 ) but this was not significant for meningeal disease caused by non-Beijing isolates ( control vs . TBM non-Beijing OR = 1 . 25 , [95% C . I . 0 . 86–1 . 82] , P = 0 . 243 ) . The strongest allelic association was between TLR2 T597C and TBM caused by Beijing genotype isolates ( control vs . TBM East Asian/Beijing; OR = 1 . 91 [95% C . I . = 1 . 28–2 . 86] , P = 0 . 001 ) . On genotypic analysis this association was also highly significant ( χ2 = 16 . 39 , P = 0 . 0003 ) ( Table 4 ) . We previously used a likelihood ratio test with Bayesian Information Criterion values to determine that the association between TLR2 T597C genotypes and TB showed best fit with a dominant ( comparing 597TT/TC vs . 597CC ) rather than a recessive ( comparing 597TT vs 597TC/CC ) model [20] . When we analyzed the association of TB caused by the Beijing lineage and TLR2 T597C using a dominant model for all types of clinical TB , we found a highly significant association ( Table 5 ) ( control vs . all East Asian/Beijing isolates: OR = 3 . 07 [95% C . I . 1 . 51–6 . 23] , P = 0 . 001] . By comparison , there was no significant association between TLR2 T597C and TB aused by non-Beijing strains ( control vs . all non-Beijing isolates: OR 1 . 75 ( 95% CI 0 . 86–3 . 56 , P = 0 . 118 ) . The association between TLR2 T597C and the Beijing strains was strongest for patients with meningeal TB ( control vs . TBM East-Asian Beijing OR = 4 . 48 [95% C . I . 2 . 00–10 . 04] , P<0 . 001 ) . Together , these results suggest that the association of SNP TLR2 T597C with TBM is strongest among those infected with the Beijing lineage .
The influence of bacterial and host genotype on the development of different forms of TB has been difficult to study in humans . We have compared bacterial and host genotype , and their interaction , across two large groups of Vietnamese adults with pulmonary or meningeal tuberculosis . The study demonstrated a relationship between M . tuberculosis phylogenetic lineage and disease phenotype: disease caused by the Euro-American lineage was significantly more likely to be pulmonary than meningeal , which suggests that this lineage may be less capable of extra-pulmonary dissemination in the study population . However , the proportion of Euro-American isolates in this study population is relatively small and therefore a larger study is required to confirm this finding . It is possible that the predominance of young males among the TBM cases presented a skewed distribution of M . tuberculosis lineages or that TBM susceptibility factors differ among the elderly or young children . It is tempting to speculate that the associations between bacterial lineage and disease phenotype are explained by the presence or absence of a functional pks 15/1 . Recent studies have suggested that the phenolic glycolipid ( PGL ) produced by some pks 15/1 intact isolates specifically inhibits the innate immune response and may be responsible for a propensity to dissemination [29] , [35] . In these studies , production of pro-inflammatory cytokines from M . tuberculosis-infected macrophages was inhibited by PGL in a dose-dependent manner . In addition , bacteria producing PGL were more capable of dissemination from the brain to other organs in animal models than others [35] . Isolates unable to express PGL – such as the Euro-American lineages - may conversely cause less extra-pulmonary disease . However , the explanation for our findings is unlikely to be as simple and extrapolation from such model studies is highly speculative . It is becoming increasingly clear that antigenic variation in M . tuberculosis is greater than previously thought and the causative mechanism of phenotypic disease variation is unlikely to be a single antigen ‘switch’ . PGL synthesis is under complex regulation and cannot be predicted simply by the presence of an intact pks 15/1 gene sequence [44] . We found no differential association with disease phenotype between the East Asian and Indo-Oceanic Lineages , although it is probable the indo-oceanic isolates do not express the PGL [44] . Of note , the patients infected with Euro-American isolates had lower mortality from TBM compared with patients infected with other lineages . This correlates well with evidence from animal models which showed rabbits infected with these strains had less severe clinical manifestations , milder focal meningeal inflammation and minimal infiltrate despite the presence of significant bacillary loads [35] . The lower mortality in human disease provides further evidence that bacterial genotype may have a significant influence on disease phenotype which could have direct clinical relevance . Bacterial genotyping may allow clinicians to identify those more likely to respond poorly to treatment in which more aggressive treatment might be beneficial . However , the number of TBM patients infected with Euro-American isolates in this study was small and a larger study is required to confirm these findings and examine potential confounders such as BCG vaccination status , immunosuppressive co-morbidities etc . Recent studies have indicated that the different lineages of M . tuberculosis are strongly associated with specific geographical regions [24] . A global phylogeography of M . tuberculosis has been proposed which suggests lineages may have become specifically adapted to their populations . Such co-evolution , or its absence , may influence disease expression and indicates interactions between bacterial and host genotype should be studied . We hypothesized that polymorphisms in genes responsible for the innate immune response to infection may influence the host response to infection and may result in increased susceptibility to disease from some bacterial lineages but not others . We found that a polymorphism in the TLR2 gene was associated with disease caused by the East Asian or Beijing lineage . This is the first time a relationship between bacterial and host genotype has been observed in TB , although it has previously been observed with other pathogens[45] . TLR2 is a trans-membrane protein which recognizes bacterial ligands - such as the 19kDa lipoprotein - and initiates a signal transduction cascade which activates dendritic cells and macrophages . The SNP T597C is a synonymous SNP that is not known to affect gene function , although we have previously demonstrated it was associated with TBM disease severity and the co-existence of miliary tuberculosis , the most extreme form of disseminated tuberculosis [20] . This suggests a polymorphism , or polymorphisms in linkage disequilibrium ( LD ) with TLR2 597C are important in multiple-facets of tuberculosis susceptibility . The causal polymorphism may lie in the promoter region , a regulatory region , or in a nearby gene , and must be identified before its effect on disease pathogenesis and interaction with Beijing genotype strains can be understood . However , it is possible that the causal mutation that is in LD with TLR2 597C may be associated with an impaired immune response to M tuberculosis and lead to more aggressive disease , prolonged bacteraemia , and an increased chance of seeding to the meninges . The Beijing genotype may further exploit the host susceptibility to infection through its own ability to subvert the host innate immune response . We have previously demonstrated a strong association between Beijing genotype and TBM in HIV positive patients in the same population [46] supporting the hypothesis that infection of an immune suppressed host with an immune subversive bacteria represent a synergistic combination that results in an increased likelihood of disease . There was no overall association of Beijing genotype with TBM in this HIV negative Vietnamese study population , although the proportion of Bejing genotype isolates was greater in the meningeal group ( 43 . 3% [81/187] of TBM isolates vs . 37 . 1% [87/234] pulmonary isolates ) , this was not significant ( P = 0 . 20 ) . Studies in other ethnicities have shown an association of Beijing genotype with extra-pulmonay disesase [36] and it remains possible that a larger study would show an association too small to reach significance here . In summary , this study provides evidence that M . tuberculosis genotype influences disease phenotype . In addition , although many reports describe host susceptibility or bacterial genetic associations with clinical phenotype in isolation , we have reported the first association between host and bacterial genotype in concert in M . tuberculosis disease . Studies of host susceptibility or pathogen virulence should be conducted in the context of both . Future vaccine candidates may need to be evaluated against a range of M . tuberculosis genotypes and host ethnicities if they are to prove globally effective , particularly against disseminated disease .
The patients were recruited to the study as previously described [19] , [20] . Briefly , patients with TBM were recruited at Pham Ngoc Thach Hospital for Tuberculosis and Lung Diseases ( PNT ) and the Hospital for Tropical Diseases ( HTD ) in Ho Chi Minh City , Vietnam between March 2000 and April 2003 . To enter the study patients had to have clinical evidence of meningitis ( nuchal rigidity and abnormal CSF parameters ) and M . tuberculosis cultured from the CSF , and be >15 years old with a negative HIV test . All patients were followed for 9 months after the start of treatment; disability was assessed in survivors by the modified Rankin score [38] . Adult patients with uncomplicated pulmonary tuberculosis were recruited between September 2003 and December 2004 at 5 district tuberculosis units ( DTUs ) from Ho Chi Minh City and the surrounding districts , chosen to represent the geographic distribution of isolates among TBM patients in order to avoid an urban/rural bias in one sample set . Cases were defined by the culture of M . tuberculosis from sputum , a chest X-ray appearance consistent with active tuberculosis without evidence of miliary or extra-pulmonary tuberculosis , and no clinical evidence of extra-pulmonary disease . As far as possible , patients were prospectively matched to TBM patients by age ( +/−5 years ) and district of residence , defined in five groups as: urban , sub-urban , rural ( surrounding HCMC ) , rural south-East or rural South-West . Matched patients were recruited from a DTU within each of these districts . Gender matching was attempted but not achieved due to a larger number of men with pulmonary TB attending the DTUs . The control group comprised of 389 DNA samples extracted from the umbilical cord blood of newborn babies born at Hung Vuong Hospital , Ho Chi Minh City , in 2003 . All samples came from unrelated individuals who were ethnic Vietnamese Kinh , as assessed by questionnaire . Written informed consent was obtained from each patient or an accompanying relative if the patient could not provide consent . All protocols were approved by ethical review committees at the HTD , PNT Hospital for Tuberculosis and Lung Disease , Hung Vuong Hospital and Health Services of Ho Chi Minh City in Vietnam . Ethical approval was also granted by Oxfordshire Clinical Research Ethics Committee UK , Oxford Tropical Research Ethics Committee UK , The University of Washington USA and the Western Institutional Review Board USA . Host genotyping and identification of TLR2 and TIRAP SNPs have been reported in detail previously [19] , [20] . Briefly , polymorphisms in both genes were identified by sequencing a randomly selected sub-group of patients with TBM . All subjects were then genotyped for the designated SNPs by an allele-specific primer extension assay ( MassARRAY™ , Sequenom , San Diego , USA ) . Analysis was performed with Bionumerics software ( Applied Maths , Sint-Martens Latern , Belgium ) and STATA 8 ( Texas , USA ) . Spoligotyping neighbour joining phylogenetic trees were created with eucldian distance coefficient on Bionumerics software . RFLP phylogenetic trees were created with 2% position tolerance and 1% optimization using Unweighted Pair Group Analysis ( UPGMA ) , dice coefficient on Bionumerics software . MIRU trees were created using UPGMA , categorical multistate coefficient . For all methods , isolates were considered clustered if 100% similarity was observed . The prevalence of genotypes among meningeal and pulmonary isolates was compared by Chi-square test . The association of LSP genotype and disease phenotype was further analysed by forward stepwise logistic regression model ( P of <0 . 05 to enter; P of >0 . 055 to remove ) to identify variables associated with disease phenotype on multivariate analysis . The variables examined in the model were LSP genotype , site of TB , age , sex and residential district . For analysis of host polymorphisms , allelic and genotypic frequencies were compared between the groups using a Chi square test . We also analyzed the data with recessive and dominant models as previously described [20] . P values of ≤0 . 05 were considered statistically significant .
|
Tuberculosis , caused by the bacterium Mycobacterium tuberculosis , kills over 2 million people each year . It is estimated that approximately one-third of the world population is infected with M . tuberculosis , though the majority will never develop active disease . The most severe form of tuberculosis occurs when the bacterium spreads to the brain to cause meningitis . We examined whether the genetic variation of the person and the bacteria influenced the type of disease a person develops . We have previously shown that certain mutations in genes of the human immune system can predispose adults in Vietnam to developing tuberculous meningitis . In this study we show that some strains of M . tuberculosis commonly found in Europe and America are less likely to cause tuberculous meningitis in Vietnamese adults than strains predominantly found in Asia . We then looked at the interaction between M . tuberculosis strains and mutations in human immune genes and show that a particular mutation , TLR2 T597C , is more commonly found in patients infected with the East-Asian/Beijing strains of M . tuberculosis . This is the first study to look at both the host and pathogen genotypes together in tuberculosis infection , and the findings suggest that the outcome of exposure to M . tuberculosis can depend on both the human genotype and the bacterial genotype .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"neurological",
"disorders/infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"microbiology/innate",
"immunity",
"immunology/innate",
"immunity",
"immunology/genetics",
"of",
"the",
"immune",
"system",
"infectious",
"diseases/bacterial",
"infections",
"microbiology/medical",
"microbiology"
] |
2008
|
The Influence of Host and Bacterial Genotype on the Development of Disseminated Disease with Mycobacterium tuberculosis
|
Schistosomiasis control mainly relies on preventive chemotherapy with praziquantel ( PZQ ) distributed through mass drug administration . With a target of 260 million treatments yearly , reliably assessing and monitoring efficacy is all-important . Recommendations for treatment and control of schistosomiasis are supported by systematic reviews and meta-analyses of aggregated data , which however also point to limitations due to heterogeneity in trial design , analyses and reporting . Some such limitations could be corrected through access to individual participant-level data ( IPD ) , which facilitates standardised analyses . A systematic literature review was conducted to identify antischistosomal drug efficacy studies performed since 2000; including electronic searches of the Cochrane Infectious Diseases Group specialised register and the Cochrane Library , PubMed , CENTRAL and Embase; complemented with a manual search for articles listed in past reviews . Antischistosomal treatment studies with assessment of outcome within 60 days post-treatment were eligible . Meta-data , i . e . study-level characteristics ( Schistosoma species , number of patients , drug administered , country , etc . ) and efficacy parameters were extracted from published documents to evaluate the scope of an individual-level data sharing platform . Out of 914 documents screened , 90 studies from 26 countries were included , enrolling 20 , 517 participants infected with Schistosoma spp . and treated with different PZQ regimens or other drugs . Methodologies varied in terms of diagnostic approaches ( number of samples and test repeats ) , time of outcome assessment , and outcome measure ( cure rate or egg reduction rate , as an arithmetic or geometric mean ) , making direct comparison of published data difficult . This review describes the landscape of schistosomiasis clinical research . The volume of data and the methodological and reporting heterogeneity identified all indicate that there is scope for an individual participant-level database , to allow for standardised analyses .
Despite its heavy burden worldwide , schistosomiasis remains a neglected tropical disease ( NTD ) which fails to attract enough attention from research and funding bodies to generate novel drugs and diagnostics [1 , 2] . Currently , only one drug , praziquantel ( PZQ ) , is available to treat schistosomiasis [3] . The World Health Organization ( WHO ) strategic plan for schistosomiasis control aims at controlling morbidity due to schistosomiasis by 2020; eliminating schistosomiasis as a public health problem by 2025; and interrupting transmission of schistosomiasis in selected countries in Africa and all other countries in the rest of the world by 2025 [4] . To reduce morbidity and transmission , the WHO recommends the use of PZQ , given as preventive chemotherapy to endemic populations [3] . Along with evidence from pre-clinical studies [5 , 6] and individual clinical trials , the efficacy and safety of PZQ is further supported by systematic reviews and meta-analyses of clinical trials [7 , 8] . However , aggregated data meta-analysis suffers from the heterogeneity in the design and reporting of the published studies [9 , 10] . This lack of standardisation is a common problem across several NTDs; specifically , in the case of schistosomiasis , it concerns diagnosis and test of cure ( diagnostic tests tend to be performed on variable numbers of samples per individual ) , outcome measures ( cure rate ( CR ) , or egg reduction rate ( ERR ) calculated by arithmetic or geometric mean ) , and time of assessment [11] . Access to detailed , individual-level participant data ( IPD ) for meta-analyses would circumvent some of those limitations and allow harmonisation of relevant variables and analyses [9 , 10] , thus enabling more meaningful analyses and comparisons across individuals or studies . This could unravel yet unknown aspects of drug efficacy and safety , in specific populations and over time . Such a model for systematic review and secondary use of clinical trials’ data has brought valuable evidence for other disease areas , whether non-communicable ( as reviewed in [10] ) or infectious ( e . g . malaria [12] ) . Monitoring PZQ efficacy is crucial , considering the size of the problem and the inherent risks of resistance [13] . In 2013 , of the some 260 million people requiring treatment for schistosomiasis ( of whom , 46% school-aged children ) , 13% did receive preventive chemotherapy with PZQ . While still far from the intended programme target , this is close to 40 million treatments ( covering two thirds of the school-aged children at need ) [14] . Recognising the potential benefits of a shared IPD database , the UNICEF/UNDP/World Bank/WHO Special Programme on Research and Training in Tropical Diseases ( TDR ) and the WHO programme for NTDs are currently working with a range of stakeholders to build a dedicated , sustainable data-sharing platform for antischistosomal efficacy studies [15] . This initiative also aims to respond to growing requirements for researchers to share their data , coming notably from funding bodies , journal editors , WHO and to some extent , the public [16–21] . The present work was performed to lay the groundwork for the development of the schistosomiasis data-sharing platform , and had three main objectives . Firstly , we sought to identify efficacy studies of antischistosomal drugs performed since 2000 , and published between January 2001 and May 2015 ( either as an article or as a conference abstract ) . We followed PRISMA guidelines [22] ( S1 Checklist ) , so that the resulting list of studies could form the basis for future collection and meta-analyses of IPD . Focus on recent studies is for practical reasons ( datasets less likely lost , corrupted or unreadable ) , but we would encourage investigators to share data from earlier studies too , if available . Secondly , we aimed to assess the volume of such IPD and its inherent characteristics , e . g . species , treatment , participants’ age-group , and study site . Such an assessment gives an overview of the expected range of data to be potentially assembled , and can help design the standardised database by highlighting key variables and meta-data to consider in its structure . Thirdly , we investigated further the heterogeneity in study design and methodologies that has already been highlighted by past systematic reviews and aggregated-data meta-analyses [11] .
To identify relevant references , the following journal databases were interrogated: the Cochrane Infectious Diseases Group specialised register and the Cochrane Library , PubMed , CENTRAL and Embase . Bibliographic references of previous meta-analyses of antischistosomal trials [7 , 8 , 11 , 23 , 24] were also manually examined . No restriction on publication language was used . Where possible , search results were filtered for ‘human’ as the species of study . Keywords belonged to three main themes reflecting the purpose of the search: ( 1 ) schistosomiasis [ ( 1 . 1 . ) schisto*; ( 1 . 2 . ) ‘schistosoma hematobium’ OR ‘schistosomiasis haematobia’ OR ‘schistosom* haematobi*‘ OR ‘urinary schistosom*’ OR ‘urogenital schistosom*’; ( 1 . 3 . ) ‘schistosoma japonicum’ OR ‘schistosomiasis japonica’ OR ‘schistosom* japonic*’ OR ‘oriental schistosom*‘; ( 1 . 4 . ) ‘schistosoma mansoni’ OR ‘schistosomiasis mansoni’ OR ‘schistosom* mansoni’ OR ‘intestinal schistosom*’ OR ‘bilharz*’]; ( 2 ) clinical study [treatment OR trial OR random* OR double-blind* OR cohort OR prospective]; and ( 3 ) specific drugs [ ( 3 . 1 . ) praziquantel OR artesunate OR artemether OR metrifonate OR albendazole; ( 3 . 2 . ) oxamniquine OR mefloquine] . Of note , although albendazole has not shown efficacy against Schistosoma spp . , it was included in the keywords for search in order to better identify studies performed by the soil-transmitted helminthiases ( STHs ) community which may have looked at schistosomiasis aside of their major study . Four subsequent searches were performed between 25 May 2015 and 12 June 2015 . The three first focussed on one major pathogenic species of Schistosoma each ( S . haematobium , S . japonicum and S . mansoni ) , and thus used keywords ( 1 . 2 . , 1 . 3 . or 1 . 4 . ) AND ( 2 ) AND ( 3 . 1 . ) . As they constituted an update of previous ( published ) Cochrane systematic searches , the two searches for literature on S . haematobium and S . mansoni were limited to records published from 2000 up to the date of search ( 25 May 2015 ) . The search for S . japonicum used no such restriction on publication date , as no prior Cochrane search had been performed . The fourth , complementary search looked at all species altogether and was restricted to publications from 2001 , but focussed on different drugs that were missing in the initial searches , therefore combining keywords ( 1 . 1 . ) AND ( 2 ) AND ( 3 . 2 . ) . Only abstracts published from 2001 , and whenever necessary , the corresponding full texts , were eventually screened for eligibility . In line with the primary objective of estimating the size of a global database of antischistosomal efficacy data , and as the likelihood of retrieving datasets dating back before 2000 is much more limited [25] , it was decided to exclude articles reporting on studies that had been completed before 1 January 2000 . Abstracts of conferences run before 1 January 2014 were also excluded , owing to the difficulty in establishing whether earlier abstracts were duplicate records of articles published within the considered time period . Finally , all case reports or case series , as well as non-primary research studies ( meta-analyses , reviews , textbook chapters , opinion papers and comments ) were excluded . Regardless of their design and associated risk of bias , remaining studies where antischistosomal ( registered or experimental ) drugs had been administered to human subjects in an endemic setting were eligible , provided that: ( i ) at least a subset of participants was positively diagnosed for infection with at least one of the three Schistosoma species of interest , prior to receiving treatment; and ( ii ) a post-treatment diagnostic test was performed on the same participants to assess outcome within 60 days , where day 0 is the day of administration of the first ( and in most cases , single ) treatment dose . The WHO recommends assessing the outcome within 21 days of therapy , in order to distinguish between cases of treatment failure and cases of early reinfection [26] . However , the maturation time of Schistosoma worms in the human host induces a delay of at least 4–5 weeks between the moment of infection and the excretion of eggs [27] . Within the 60 days post-treatment window , individuals who were initially cured but got re-infected are still likely to excrete low numbers of eggs ( if at all ) , and to appear negative upon diagnosis by egg count . The 60-day limit was thus retained , to avoid excluding too many efficacy studies which performed an outcome assessment beyond the recommended 21-day timeframe . Relevant information from eligible studies published in English or French was extracted by one researcher , using a detailed variable dictionary to facilitate consistent extraction . Help from co-authors and from native Chinese speakers was sought , as required . Extracted information was recorded using a spreadsheet document ( Microsoft Excel 2000 ) , and analysed using R ( version 3 . 1 . 3 ) [28] . The source data used for analysis , the associated dictionary and the R code are all available as supporting documentation in S1 Dataset ( Zip ) . Some of the eligible studies had a broad scope , and were looking simultaneously at schistosomiasis and at other parasitic infections , in particular STHs . Articles on such studies did not systematically report on the exact number of participants with a Schistosoma infection at baseline , thus the need to estimate this number for the purpose of this review . Whenever possible , the number of Schistosoma-positive participants was re-calculated from the specified prevalence of infection among enrolled participants . If the prevalence amongst enrolled participants was unknown , the reported prevalence in the study area was used instead . Finally , in the rare case when there was no mention of prevalence at all , half of the cohort was arbitrarily considered infected with Schistosoma spp . at baseline ( i . e . assumed prevalence of 50% among enrolled participants ) . Of note and unless more specific information was provided in the article , the retained frequency of infection ( exact prevalence in study population , reported prevalence in study area , or 0 . 5 ) was then consistently applied as a correction factor when estimating the number of relevant participants in subgroups ( e . g . treatment group , group followed-up at later time-points ) : this implies that the subgroup characteristic and the initial infection status of the participant would be independent . After extraction and coding of all information as per the variable dictionary , the datasets were checked for potential errors . To do so , each article was systematically re-examined for variables considered most ‘impactful’ . These were primarily the variables recording numbers of relevant participants ( Schistosoma-positive at baseline , and who successfully received the assigned treatment regimen ) at both enrolment and follow-up , and in each assignment arm; because those data directly impact on the total number of IPD , whose estimation was our primary objective . The species and drug ( s ) under study were also double-checked , as the information coded in those variables was used to filter articles by category ( S . haematobium vs . S . mansoni , PZQ vs . other drugs , etc . ) and to create specific datasets for analyses by sub-category . Finally , impactful variables also included the time-points of post-treatment assessment of outcome , as they influenced inclusion/exclusion of the article from the analysis . This double-checking led to the correction of 3 to 12 errors per variable , mainly due to ambiguity in the original articles ( e . g . unclear flowchart of included , treated and followed-up participants ) , and their correction did not significantly change the total of estimated IPD . A subset of articles was also randomly selected for full , independent re-analysis . This step unravelled 3 typos , which , spread across a total of 45 variables ( excluding the core variables previously double-checked ) and 12 articles , suggest an error rate of 0 . 56% in the data as a whole , and of >1% per variable .
The three species-specific searches of the literature in electronic resources yielded 545 , 440 and 272 study abstracts for S . mansoni , S . haematobium and S . japonicum , respectively . The later search for additional drugs gave only 11 additional results , of which 6 abstracts reported on studies of S . mansoni , and 5 on S . haematobium . In parallel , 190 references were identified in the bibliography of published meta-analyses [7 , 8 , 11 , 23 , 24] . Overall , those amounted to 914 unique abstracts published since 1 January 2001 , of which 90 studies were retained after eligibility screening ( Fig 1 ) . Although most studies could be thoroughly evaluated , the case of three abstracts [29–31] remained unclear , and we could not access the corresponding article . In addition , the eligibility of two other studies [32 , 33] could not be ascertained despite access to full text . At this stage , and as this review is a preliminary phase prior to possible collection of IPD and performance of clinically relevant meta-analyses , no attempt was made to contact the authors and clarify missing information . This step shall be performed later , along with a public call addressed to the schistosomiasis community to help identify more studies that might have been missed by this scoping review ( e . g . studies that were not published , or published in non-indexed journals ) . The 90 included studies correspond to 104 cohorts , where a ‘cohort’ is the largest possible subset of study participants from the same country , which followed the same protocol and thus share the same study-level characteristics ( except for the assigned intervention in multi-arm studies ) . Most studies ( 79 ) involved a single cohort , but some looked at different groups such as infants and adults who were followed up at different time-points , or involved sites across several countries with different parasite species and diagnostic approaches . Those study groups were thus considered to be separate cohorts , and the analysis by cohort is preferred whenever comparing the specific methodologies and protocols applied . The 90 studies enrolled an estimated total of 20 , 517 participants of relevance for schistosomiasis treatment efficacy meta-analyses . One conference abstract did not provide any information to estimate the number of participants: accordingly , this record did not contribute to analyses by number of participants , but was still considered for analyses by cohort or study . Throughout this article , whenever mentioning a number of participants , it is derived from the number of enrolled participants infected with Schistosoma spp . at baseline , and who received the full treatment regimen they were assigned . The number of participants refers to the initial cohorts , as losses to follow-up were inconsistently reported and therefore could not be accounted for here ( estimates suggest that the total of participants with complete data is about 17 , 500 ) . A major advantage of statistical meta-analyses based on IPD pooled from multiple studies is to provide large numbers . With this method , specific sub-populations whose outcomes are usually ‘hidden’ in aggregated data could be individually considered for eligibility in re-analyses on a case-by-case basis , depending on the objectives . Here for instance , 26 cohorts are formed of participants infected with several possible parasites , which made it difficult to extract information on Schistosoma-infected participants using published information only ( Fig 2 ) . Depending on the purpose of the meta-analysis , it could be interesting to retrieve data about these estimated 4 , 916 participants infected with schistosomiasis , and treated with drugs having antischistosomal activity as part of larger studies focussing on other parasites ( Fig 2 ) . The 104 cohorts were also recruited in varied endemic settings and countries , and tested several therapies—mostly various doses and regimens of PZQ , but also a dozen other therapies ( Fig 3 ) . Considering the large amount of participants having received PZQ at the dose of 40 mg/kg of body weight ( the WHO-recommended regimen for preventive chemotherapy [3] ) , pooled data could help better characterize responses and identify subgroups with sub-optimal responses , and evaluate efficacy over time and in different settings ( the 104 cohorts were recruited in 26 different countries ( Fig 4 ) and in various contexts ( Fig 5 ) ) . In addition , the 20 , 517 IPD are likely to represent different sub-populations of interest and that remain under-studied , such as preschool-aged children ( 30 of the 104 cohorts clearly recruited at least one participant aged strictly less than 6 years ) , or pregnant women ( 2 cohorts focussed on this group , i . e . a total of 452 participants and possibly more , as several cohorts seemed not to specifically exclude those women ) . Most cohorts were relatively small , with 62 . 5% of them involving less than 200 participants with schistosomiasis at baseline ( Fig 6 ) , thus the need to pool data together to reach the sample size giving enough statistical power to unravel drug effects in those smaller populations . Not all of the studies included in the present review would meet more stringent eligibility criteria as required for inclusion in Cochrane systematic reviews [36] . No judgement on the type of study design was performed prior to inclusion . Potentially all good-quality drug efficacy data , even if not obtained from a randomised controlled trial ( RCT ) , can contribute to IPD meta-analyses ( comparison of continued drug efficacy in specific areas , modelling and model validation , trial methodology questions , etc . ) , and thus should be eligible for inclusion in a global database . Less than half ( 41 ) of the studies included here were comparative , and a third of them ( 30 studies ) delivered a comparator intervention to the control arm ( Fig 7 ) . Of these comparative studies , 26 studies were randomised , out of which 12 mentioned a computer-generated randomisation sequence ( Fig 7 ) . Access to IPD would facilitate the appraisal of study design , for example by enabling to check the reliability of the randomisation method [9 , 10] . Owing to those loose eligibility criteria regarding study design , the studies included in the present review did not necessarily have the assessment of efficacy as their primary objective . Some rather focussed on specific immune responses following drug administration ( 15 studies ) , or on the search for specific biomarkers of infection and new approaches to diagnosis ( 14 studies ) . This means that not all studies are expected to report a drug efficacy outcome such as a CR or an ERR . The CR , or proportion of participants who were infected at baseline and negative at follow-up [37] , was the most frequently reported measure of drug efficacy , with 66 of the 90 studies providing such information ( Fig 8 ) . As the diagnosis of schistosomiasis is mainly performed by counting Schistosoma eggs in excreta ( urine or stool for urinary and intestinal schistosomiasis , respectively ) , drug efficacy can also be assessed as ERR , or variation in the mean number of eggs found in the excreta of the population before and after treatment [26] . Studies however tend to compute the ERR in different manners , and in particular , to derive it from an arithmetic or a geometric mean ( here 7 and 39 articles , respectively , Fig 8 ) . Such variability in efficacy reporting is a major issue in meta-analysis of aggregated data , which could be solved by recalculating measures using the original IPD , and ideally , using the original raw egg count from each of several measurements done for each participant . In all cohorts , the primary method to diagnose infection at screening was egg detection and counting in excreta . Eight cohorts also diagnosed participants by quantifying levels of schistosomiasis circulating anodic or cathodic antigen in body fluids ( urine or serum ) . Of those 8 cohorts , 3 used only the latter circulating antigen method to assess the outcome after treatment . The most frequent method to diagnose infection with Schistosoma spp . is thus counting eggs; however , differences exist in how samples are prepared ( different techniques may be used to process samples prior to egg count under microscopy , especially for the preparation of stools , Fig 9 ) , and how many samples are analysed ( numbers of samples and/or test repeats , Fig 9 ) . Since data are reported as means , and the sensitivity of the method depends on the number of independent egg counts for each subject , access to pooled IPD enables better to quantify the extent of this variation in sensitivity , and how it impacts on reported treatment outcomes [24] . The time at which outcome is assessed post-treatment also varies across studies ( Fig 10 ) : 48% of the cohorts ( 50 cohorts corresponding to 8 , 144 participants ) have at least one data point at 3 to 4 weeks post-treatment , and 20 of those cohorts also have later assessments , sometimes much beyond the analysed 60-day limit . Such cohorts with data points at both 3–4 weeks and later provide opportunities for analyses distinguishing between initial cure and potential reinfection patterns .
We are cognisant that not all the data identified here may be available ( e . g . , electronic dataset non-existent , corrupted or lost ) , or be made available ( some data generators might be reluctant to share ) [25 , 38] , but also that more datasets might be available than those captured here . Those include national control programmes surveys ( most of which are generally not published ) , and studies still on-going and/or awaiting publication . The delay between study completion and publication is sometimes very long , ranging from 5 months to 8 years in the case of the articles analysed here ( S1 Fig ) ; and only 12 of 90 studies were found to be registered in ClinicalTrials . gov or Controlled-Trials . com . In addition , several studies published locally in languages other than English could also have been missed . No restriction on language was used in the search , but regional registries such as LILACS ( South American journals ) [39] and CNKI ( Chinese journals ) [40] were not interrogated . Only 4 of the 90 included articles were in French , and none was in Portuguese or Chinese despite Brazil and China having reputed control programmes and productive research institutes . Few to no data on oxamniquine ( OXA ) and metrifonate ( MTF ) could be identified in the post-2000 timeframe . A search of PubMed articles published in the 1986–2000 period identified at least 2000 and 500 more IPD for OXA and MTF , respectively . Though experience suggests that such ‘older’ data will be more difficult to retrieve [25] , efforts to collect past data from Brazil on OXA are under way and should be promoted , in order to provide a more complete picture of the efficacy of several antischistosomal drugs , and its variations over time . Similarly , the collection of data from China should yield additional data on S . japonicum , which is currently under-represented ( 2 , 327 IPD ) as compared to S . haematobium ( 9 , 833 IPD ) and S . mansoni ( 10 , 790 IPD ) . Moreover , local variations in the preferred post-treatment time of outcome assessment may have led to the exclusion of yet relevant studies . When consulted on the development of an antischistosomal data-sharing platform , some investigators suggested to extend the boundary of acceptable time of first assessment to 3 months ( 90 days ) , instead of the 2 months ( 60 days ) used as a cut-off for inclusion in the present review: 104 articles were excluded at screening on the grounds of the outcome assessment occurring too late , of which many had performed it at 3 months post-treatment . The preferred timeframe for study inclusion should also depend on the purpose of the subsequent meta-analysis , whether strictly restricted to drug efficacy assessment , or aiming at estimating long-term effects on infection intensity in the population . Despite its limitations , the present review confirms and extends the problem of the variability in clinical research methods and reporting [11] , which makes data summaries difficult to re-analyse and undermines the strength of evidence . Some sources of variability could be corrected with access to IPD and application of standardised analyses , as demonstrated by the example of efficacy outcomes: with IPD , it is possible to recalculate ERRs from distinct studies in a homogeneous manner using arithmetic means ( as recommended by the WHO [26] ) , and to explore and compare alternative ways of expressing efficacy , as already done on an initial dataset of up to 4 , 740 individuals [24 , 41] . Other sources of variability will be more difficult to correct , but a global database would allow to pool data from studies with consistent diagnostic approaches or , as shown in Olliaro et al . [24] , to analyse the entire database using the WHO standard operating procedures ( one single test of one single sample [26] ) . Based on the present analysis and on our past experience of performing IPD meta-analyses [24 , 41] , we find that the following variables would be most useful: site ( s ) and country of study , Schistosoma species under study , year of study start and stop , diagnostic method ( number of samples and test repeats , technique of samples’ preparation ) , regimen ( drug name , dose , formulation and treatment regimen , if applicable ) and time of follow-up ( s ) . Dates of different events ( recruitment , treatment and follow-up ( s ) ) and associated outcomes ( e . g . diagnostic result , adverse events ) should also be collected for each study participant . As a final consideration , although 20 , 000 ( and possibly more ) IPD is significant and justifies efforts for the creation of an IPD shared database , this amount of data for all the published schistosomiasis treatment trials of the past 15 years is disappointingly low , and reflects the lack of investments in clinical research in this aptly-named NTD . The present scoping exercise supports the interest of an antischistosomal efficacy data-sharing platform , as more than 20 , 000 IPD could contribute to extensive meta-analyses , using a common analytical plan . This would improve our knowledge of how effective treatments are in the overall population as well as in subgroups ( age , location , etc . ) and over time; and of how to analyse data in a meaningful way by exploring and comparing different approaches . The need for such a database is further emphasised by the push from funders and journals to make research data more widely accessible and optimally used . With this in mind , schistosomiasis stakeholders , researchers and policymakers have been invited to discuss their views on this matter at a WHO-convened meeting held in September 2015 [15] . Providing that a clear data governance structure is defined , especially regarding how data which are shared will be accessed and used , and that measures are taken to address the need for data management and analysis capacity-building , there was general support to the establishment of such a database towards producing a stronger evidence-base for schistosomiasis treatment .
|
Schistosomiasis is a parasitic disease prevalent in tropical climates , especially Africa . To control morbidity and eventually eliminate the disease , the World Health Organisation recommends that populations living in endemic areas be preventively treated with praziquantel . This means hundreds of million treatments over time , which calls for careful monitoring of drug efficacy . Meta-analyses of aggregated data from clinical trials support this policy , but conclusions are limited because studies use diverse methodologies , and detailed analyses ( e . g . by age-groups ) are not possible . Access to primary data generated by trials would strengthen the evidence and allow answering key questions , such as what the optimum paediatric dose is . This review assesses the number of trials conducted on antischistosomal drugs and their characteristics ( where studies were conducted , in which population , what drug ( s ) /dose ( s ) were administered , and the methodologies applied ) , to determine the merit and scope of a common database of antischistosomal efficacy data .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"medicine",
"and",
"health",
"sciences",
"helminths",
"tropical",
"diseases",
"parasitic",
"diseases",
"animals",
"mathematics",
"statistics",
"(mathematics)",
"pharmaceutics",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"schistosoma",
"japonicum",
"schistosoma",
"haematobium",
"mathematical",
"and",
"statistical",
"techniques",
"helminth",
"infections",
"schistosomiasis",
"meta-analysis",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"drug",
"therapy",
"statistical",
"methods",
"organisms"
] |
2016
|
The Schistosomiasis Clinical Trials Landscape: A Systematic Review of Antischistosomal Treatment Efficacy Studies and a Case for Sharing Individual Participant-Level Data (IPD)
|
WHO’s Global Programme to Eliminate Lymphatic Filariasis ( LF ) uses mass drug administration ( MDA ) of anthelmintic medications to interrupt LF transmission in endemic areas . Recently , a single dose combination of ivermectin ( IVM ) , diethylcarbamazine ( DEC ) , and albendazole ( ALB ) was shown to be markedly more effective than the standard two-drug regimens ( DEC or IVM , plus ALB ) for achieving long-term clearance of microfilaremia . To provide context for the results of a large-scale , international safety trial of MDA using triple drug therapy , we searched Ovid Medline for studies published from 1985–2017 that reported adverse events ( AEs ) following treatment of LF with IVM , DEC , ALB , or any combination of these medications . Studies that reported AE rates by treatment group were included . We reviewed 162 published manuscripts , 55 of which met inclusion criteria . Among these , 34 were clinic or hospital-based clinical trials , and 21 were community-based studies . Reported AE rates varied widely . The median AE rate following DEC or IVM treatment was greater than 60% among microfilaremic participants and less than 10% in persons without microfilaremia . The most common AEs reported were fever , headache , myalgia or arthralgia , fatigue , and malaise . Mild to moderate systemic AEs related to death of microfilariae are common following LF treatment . Post-treatment AEs are transient and rarely severe or serious . Comparison of AE rates from different community studies is difficult due to inconsistent AE reporting , varied infection rates , and varied intensity of follow-up . A more uniform approach for assessing and reporting AEs in LF community treatment studies would be helpful .
Infection with the filarial nematode parasites Wuchereria bancrofti , Brugia malayi , or Brugia timori is known as lymphatic filariasis ( LF ) . These infections cause severe , disabling conditions including lymphedema , elephantiasis , and hydroceles in tens of millions of people in tropical and subtropical countries . Annual mass drug administration ( MDA ) coordinated by WHO’s Global Programme to Eliminate LF ( GPELF ) has significantly reduced LF transmission in many of the 78 initially endemic nations [1–3] . Yet LF remains far too common , with tens of millions infected and 850 million people at risk of acquiring the infection in 53 countries [3] . With approximately 500 million people receiving MDA for LF each year , understanding , anticipating , and preparing the targeted population for MDA-related adverse events ( AEs ) is important for program success . Medications used for MDA include diethylcarbamazine ( DEC ) , ivermectin ( IVM ) and albendazole ( ALB ) . The combination of IVM plus ALB is used in areas of Africa where onchocerciasis ( river blindness ) is co-endemic with LF . Twice yearly ALB alone is recommended for LF-endemic areas of Africa that are co-endemic for loiasis , and DEC plus ALB is used in the rest of the world . Serious ( life-threatening ) AEs due to MDA are exceedingly rare [4–7] . However , when they do occur they can profoundly impact the treated community and jeopardize program success [8] . When communities are well-informed about the type and severity of AEs to be expected , they may be less likely to avoid MDA out of fear of AEs . Furthermore , the knowledgeability of community health workers ( drug distributors ) can be a major determinant of MDA adherence [8] . A clear understanding of the nature of expected AEs should empower program managers and community health workers to prepare their communities to anticipate and accept transient AEs , which may in turn improve compliance with MDA and facilitate LF elimination efforts . A promising new combination therapy for LF that combines a single dose of IVM , DEC , and ALB ( IDA ) appears to be highly effective [9] , and its safety is has been evaluated in large community-based studies in several locations ( ClinicalTrials . gov Identifier: NCT02899936 ) [10] . This manuscript’s purpose is to provide context for understanding the safety of the new IDA treatment by reviewing published data on the rates and nature of AEs following single-dose treatment for LF with any of the IDA medications . As previously noted by many others , AE reporting in LF treatment trials is highly variable and potentially affected by multiple factors including blood microfilaria ( Mf ) counts , treatment regimens , filarial species , population demographics , and importantly , the thoroughness of post-treatment surveillance . We therefore sought to review AE data from published LF treatment studies to further understand the effect of these parameters on AE rates and severity . Our objective was to evaluate reports of AEs following single dose LF treatment of children and adults with IVM , DEC , or ALB ( either as monotherapy or in multidrug combination regimens ) , published since 1985 . In this report we first present a broad summary of the literature reviewed and then a quantitative synthesis of published AE rates from studies meeting our specified inclusion criteria .
We reviewed AE data from studies of LF treatment with single-dose regimens that were published between 1985 and 2017 . We searched Ovid Medline and Embase for any articles with Medical Subject Headings ( MeSH ) terms “Elephantiasis , Filarial” and “Drug Therapy” plus any of the following terms: “Adverse Events” , “Poisoning” , or “Toxicity” . We limited our search to English or French language manuscripts dealing with human infections . The most recent search was conducted on 21 Aug , 2017 . Two authors ( PB and CH ) reviewed each publication and gathered additional pertinent publications from articles referenced therein . Publications with sufficient AE data were selected for a quantitative analysis of AEs as described below . We did not pre-specify nor register a review protocol . We did not attempt to contact authors to identify additional studies . Studies published after 1985 that reported AE rates following single-dose LF treatment with IVM , DEC , or ALB ( alone or in combination ) were included . Studies dealing with multi-day courses ( generally of DEC ) were reviewed , but excluded from the quantitative analysis , as were studies that either provided inadequate information on AEs by treatment arm or did not conduct AE surveillance within one week following treatment . Complete inclusion and exclusion criteria are shown in Table 1 . We followed the PRISMA Statement for Reporting Systematic Reviews and Meta-Analysis [12]; the completed PRISMA checklist is available as a supplemental file ( S1 Table ) . From studies meeting inclusion criteria we extracted data including: study location ( country ) , age range and gender of participants , intensity of surveillance , treatment regimen , Mf prevalence , geometric mean Mf counts , presence of co-infections , overall rate of AEs , and rates for any specific AEs reported . For studies that reported AEs following multiple MDA treatment rounds , we included only the AE rates that occurred after the first treatment . For studies in which one but not all treatment arms met inclusion criteria ( for example , when single dose IVM or DEC was compared to 12 days of DEC ) , we included data only from the arm ( s ) meeting inclusion criteria . The number of participants reported in our analysis is the number for whom AE surveillance was conducted , which was sometimes lower than the total number treated . For example , one study conducted active post-MDA surveillance within a subset of 483 persons among 8 million people treated [18]; in our analysis , the N for this study was 483 . All extracted data were analyzed using Stata version 12 . 1 ( College Station , TX ) . Because the data were not normally distributed , we report means and interquartile ranges ( IQR ) and use boxplots for graphical representation . Since AE reporting was insufficiently uniform among included studies , we did not attempt a formal meta-analysis of AE rates , nor did we attempt statistical analyses . Rather , we sought to present a graphical synthesis of data from these disparate studies to illustrate the range of data and an estimate of central tendency ( median and interquartile range ) . To assess for reporting bias in individual studies , we stratified surveillance for AEs in each study as active ( individual participants were contacted and asked about AEs ) or passive ( individuals with AEs had to seek out the study team to report ) . The quality of active surveillance was further categorized as “high” ( at least daily contact during the first 72 hours ) , “moderate” ( at least one contact within first 72 hours ) , or “low” ( participants contacted after 72 hours ) . Although we hoped to analyze the effect of each extracted variable on reported AE rates , we found that the quality of data reported for most parameters was insufficient . We therefore limited our analysis to an ad hoc comparison of treatment regimens , Mf status , and intensity of AE surveillance .
Many informative articles that reported AEs following treatment for LF could not be included in our quantitative analysis either because they reported composite AE scores rather than rates , or because they did not report AE rates separately by treatment group . We have attempted to review some of the observations from both included and excluded publications in the following paragraphs . To summarize published rates of AEs following single-dose treatment of LF , and to explore how these might differ by treatment medication and AE surveillance , we compiled data from articles with adequate AE reporting into a combined analysis . Among 162 full text articles reviewed , 55 contained AE data that met criteria for inclusion ( Fig 1 ) . There was considerable heterogeneity in the way that AEs were reported in these studies; 34 reported both the aggregate incidence of AEs ( i . e . the number of persons experiencing any AE ) and the percentage of persons experiencing specific AEs . Seventeen studies reported an aggregate incidence but not specific events , and four reported specific events but not aggregate incidence . Methods of AE ascertainment varied widely between studies , from intensive in-hospital monitoring to passive reporting in community-based trials . For the purposes of our analyses we grouped the studies into two main types: ( 1 ) clinical trials with active AE surveillance and ( 2 ) community studies with either active or passive surveillance . The former group comprises studies in which 100% of participants were Mf positive , while community studies had varied Mf rates ( Table 2 ) .
In this review we initially sought to quantify the effects of various factors on AE rates that occur following MDA for LF . We quickly realized that the heterogeneity in the way AEs have been reported in the literature would not allow a meaningful quantitative multivariate analysis . We nevertheless felt a compilation of reported AE rates would be beneficial . Despite the limitations of combining data from methodologically disparate studies , we believe the compiled data illustrate the following main points: 1 ) AEs are very common in microfilaremic patients after single-dose treatment of LF with drugs ( IVM and DEC ) that rapidly reduce Mf counts . 2 ) AEs are much less common in amicrofilaremic participants , regardless of treatment regimen . 3 ) Passive surveillance tends to underestimate the occurrence of AEs , and 4 ) Heterogeneity in the stringency of AE surveillance and format of AE reporting makes comparisons between studies difficult . The relationship between AE rates and the prevalence of microfilaremia is illustrated by the striking differences between study arms with 100% microfilaremia and those with no microfilaremia . It would have been interesting to compare Mf prevalence to AE rates among the community studies with varied Mf prevalence; this was not attempted because of the variability in AE reporting for these studies and because uncertainty regarding true Mf rates would have made this comparison unreliable . It is clear that community proclivities for reporting AEs vary from place to place and study to study . This is perhaps most evident in reported AE rates after placebo treatment . Studies with highly active AE surveillance in Haiti and Tahiti reported high AE rates after placebo treatment [54 , 84 , 93] , but AE rates were low after placebo treatment in Ghana and India [86 , 120] . This place-to-place AE reporting variability is also evident in the wide range of AE rates reported among different studies with the same treatment regimens ( see Figs 3 and 5 ) . Potential reasons for this might include the prevalence of STH or other helminth infections , differing intensities of LF infection , and varying cultural norms . In addition , where populations have been sensitized to expect AEs following MDA , more AEs may be perceived [18] . Nearly all the studies cited in this review reported AE rates in some manner , but we were only able to include 55 in the quantitative synthesis . The primary reason for excluding studies was that they did not present AE data in a way that linked AE rates to treatment regimens . For example , many studies reported AE severity scores rather than rates . Others reported that AE rates did not differ significantly between treatment groups , but did not report the numbers for each group . When AE rates were reported by treatment group , comparisons were often hampered by non-standardized AE reporting procedures . Some authors did not report the timeframe over which AE surveillance was conducted , making it difficult to surmise whether early or late AEs may have been missed . Although most studies included in our analysis described whether AE surveillance was active or passive , many contained insufficient detail to determine how sensitive the study procedures were for detecting AEs . For example , ascertainment rates ( the proportion of participants in community-based studies who were actually visited and queried about AEs ) were almost never reported . This review has several strengths and weaknesses . The primary strength is that it compiles data from 30 years of published studies . It also illustrates how variable AE reporting can be , and it provides a context for interpreting AE rates observed in future LF treatment studies . One weakness was our inability to include data from many high quality studies that did not report AEs by treatment arm . In addition , because we restricted our analysis to studies of single-dose therapy , many rich and highly informative studies that used multi-dose treatment regimens were excluded . In general , the pattern of AEs reported in such studies was similar to single dose studies . That is to say , the rate and severity of AEs increased with increasing Mf counts and most AEs occurred during the first 48 hours after the initial treatment dose [38 , 82] . The heterogeneity in AE reporting among the reviewed studies highlights the need for a more structured approach to AE reporting in LF treatment studies . Although this problem is not unique to filariasis [128] , it can be compounded by the nature of community-based studies . We therefore suggest the following measures for improving AE reporting in community based treatment trials for LF and other neglected tropical diseases ( Box 1 ) . 1 ) Clearly specify the methods for ascertaining AEs . Indicate whether an attempt was made to contact each participant ( active surveillance ) or whether participants were required to seek out study staff to report AEs ( passive surveillance ) . Indicate when and how often participants were contacted . Avoid ambiguous language such as “followed closely” , or “closely monitored” . Rather , describe what was actually done . For example , “treated individuals were visited daily in their homes for five days after treatment” . 2 ) If surveillance was active , report the ascertainment rate; that is , the proportion of participants sought during surveillance that was actually found . Knowing what proportion of participants actually contributed to the reported AE rates will help the reader assess the reported findings . For example , one study reported , “All subjects [were] asked to come to the study site on day 2 and day 5… . In addition , team members also went door to door . ” The door to door contacting was presumably meant to ascertain AEs in subjects not reporting at the study site , since subjects may choose not to present for follow-up either because they feel well and see no need , or because they feel ill and don’t wish to leave their homes . The higher the proportion of participants for whom actual AE status is not ascertained , the greater the uncertainty regarding the reported AE rates . Unfortunately , the study cited—and most other community studies we reviewed—did not report ascertainment rates . 3 ) Report numerators and denominators . When severity scores are used ( for example , 1 = mild , 2 = moderate , 3 = severe ) to compare AEs between study groups , the actual number of persons experiencing AEs should also be reported so that rates can be calculated . The difference between one person with a severe AE and three people with one mild AE each is important , and the reporting of AE data should allow the reader to distinguish the difference . In addition to clearly specifying the number of persons experiencing AEs ( the numerator ) , the denominator should be clearly defined . In active studies , we suggest reporting AE rates as the proportion of those experiencing AEs over the number actually assessed . For example , five persons experiencing AEs among 20 patients treated should be reported as 50% ( not 25% ) if only 10 of those treated were actually assessed . 4 ) Use standardized grading criteria for reporting AE severity . Examples include the National Cancer Institute’s Common Terminology Criteria for Adverse Events ( available at https://evs . nci . nih . gov/ftp1/CTCAE/About . html ) or the Division of AIDS Table for Grading the Severity of Adult and Pediatric Adverse Events ( available at http://rsc . tech-res . com/clinical-research-sites/safety-reporting/daids-grading-tables ) . 5 ) Follow CONSORT guidelines for better reporting of harms in clinical trials [129] . In conclusion , this review has shown that AEs following single dose treatment of LF are common and should be expected in microfilaremic patients . This information provides a useful context for understanding AEs observed with new treatments for LF . Clear and detailed reporting of AEs in community treatment studies is essential to accurately inform elimination program workers and their communities , and to set appropriate expectations . The fear of MDA-associated AEs is often out of proportion to the actual risk , because most post-treatment AEs are mild and transient . A frank explanation of AEs as a marker for treatment efficacy by program managers and community health workers may improve compliance with MDA and facilitate LF elimination efforts .
|
WHO’s Global Programme to Eliminate Lymphatic Filariais ( LF ) supports annual mass drug administration to over 400 million people in LF-endemic areas each year . Two drug combinations ( either DEC or ivermectin , given with albendazole ) have been recommended in most endemic areas . With the exception of well-described serious adverse events ( AEs ) occurring in patients with high level loiasis , severe AEs due to these medications are extremely rare . Mild to moderate AEs , however , are common , particularly in patients with active filarial infection . In this manuscript we synthesize published data on AEs following single-dose treatment of LF with ivermectin , DEC , or albendazole . This provides a background against which to compare the safety of triple drug therapy ( ivermectin , DEC , and albendazole ) recently endorsed by WHO , and provides a useful context for evaluating safety of new treatments for LF . The compiled data illustrate that transient , mild to moderate AEs following single-dose LF treatment are common in microfilaremic patients and are much less common in amicrofilaremic patients . They also show that passive surveillance for post-treatment AEs underestimates AE incidence and suggest that adherence to common reporting standards would improve the usefulness of AE reporting in filariasis studies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"clinical",
"research",
"design",
"tropical",
"diseases",
"parasitic",
"diseases",
"wolbachia",
"nematode",
"infections",
"research",
"design",
"filariasis",
"signs",
"and",
"symptoms",
"pharmaceutics",
"neglected",
"tropical",
"diseases",
"bacteria",
"lymphatic",
"filariasis",
"research",
"and",
"analysis",
"methods",
"adverse",
"events",
"loiasis",
"quantitative",
"analysis",
"helminth",
"infections",
"diagnostic",
"medicine",
"fevers",
"biology",
"and",
"life",
"sciences",
"drug",
"therapy",
"organisms"
] |
2018
|
Adverse events following single dose treatment of lymphatic filariasis: Observations from a review of the literature
|
Plasmid conjugation plays a significant role in the dissemination of antibiotic resistance and pathogenicity determinants . Understanding how conjugation is regulated is important to gain insights into these features . Little is known about regulation of conjugation systems present on plasmids from Gram-positive bacteria . pLS20 is a native conjugative plasmid from the Gram-positive bacterium Bacillus subtilis . Recently the key players that repress and activate pLS20 conjugation have been identified . Here we studied in detail the molecular mechanism regulating the pLS20 conjugation genes using both in vivo and in vitro approaches . Our results show that conjugation is subject to the control of a complex genetic switch where at least three levels of regulation are integrated . The first of the three layers involves overlapping divergent promoters of different strengths regulating expression of the conjugation genes and the key transcriptional regulator RcoLS20 . The second layer involves a triple function of RcoLS20 being a repressor of the main conjugation promoter and an activator and repressor of its own promoter at low and high concentrations , respectively . The third level of regulation concerns formation of a DNA loop mediated by simultaneous binding of tetrameric RcoLS20 to two operators , one of which overlaps with the divergent promoters . The combination of these three layers of regulation in the same switch allows the main conjugation promoter to be tightly repressed during conditions unfavorable to conjugation while maintaining the sensitivity to accurately switch on the conjugation genes when appropriate conditions occur . The implications of the regulatory switch and comparison with other genetic switches involving DNA looping are discussed .
Bacteria exchange genetic material at high rates by different processes , which are collectively named Horizontal Gene Transfer ( HGT ) . HGT can be beneficial for bacteria because the newly acquired DNA may endow them with novel features enabling them to adapt to changing conditions in the environment , i . e . rapid evolution . On the other hand , HGT is notorious for its role in the dissemination of virulence/pathogenicity determinants and antibiotic resistance . The main mechanisms responsible for HGT are transformation mediated by natural competence , transduction and conjugation [1]–[6] . The latter mechanism , -conjugation- , concerns the transfer of a DNA element from a donor to a recipient cell . Conjugative elements containing all the information required for DNA transfer of a donor to a recipient cell are often found on plasmids , but they can also be embedded within a bacterial chromosome . These latter forms are generally named integrative and conjugative elements ( ICE ) . Some basic features of the conjugation process are conserved among plasmids [for review see] , [ 7]–[10] . In most cases , a single-stranded DNA ( ssDNA ) , which is generated by a rolling circle-like mode of DNA replication , is transferred into the recipient cell through a membrane-associated intercellular mating channel , named transferosome , which is a form of type IV secretion system . Conjugative plasmids can be exploited for the construction of tools to genetically modify bacteria of clinical or industrial relevance that are reluctant to genetic manipulation by other ways . Besides its intrinsic scientific interest , a detailed understanding about how conjugation genes are regulated is crucial to design strategies helping to interfere with the rapid spread of antibiotic resistance , and for the construction of genetic tools based upon conjugative plasmids . Various conjugative plasmids have been studied in considerable detail [for review see] , [ 7]–[10] . Although most of the well-studied conjugative plasmids replicate in Gram-negative bacteria , an increasing interest in conjugative plasmids of Gram-positive bacteria has resulted in the recent analysis of conjugative plasmids from for instance streptococci , enterococci , staphylococci and clostridia [11]–[17] . However , conjugation systems present on the Gram-positive soil bacterium Bacillus subtilis had not been reported until recently . This is most remarkable taking into account that ( i ) it is one of the best-studied Gram-positive bacteria; ( ii ) it has important industrial applications; and ( iii ) it is closely related to pathogenic and fastidious bacilli [for review see] , [ 18 , 19] . Moreover , several B . subtilis strains are gut commensals in animals including humans [20] . B . subtilis plasmids may therefore play an important role in HGT in different environments . We chose the B . subtilis plasmid pLS20 for our studies . Originally , this 65 kb plasmid was identified in the Bacillus subtilis natto strain IFO3335 that is used in the fermentation of soybeans to produce “natto” , a dish that is popular in South Asia [21] . Previous studies on pLS20 have shown that it is conjugative in liquid media as well as on solid media [22] , [23] . The presence of pLS20 has a broad impact on the physiology of the host , and the localization of some components of the conjugation machinery has been determined [24] , [25] . The replication region of pLS20 has been characterized , and it has been demonstrated that it uses a dedicated segregation mechanism involving the actin-like Alp7A protein [26] , [27] . pLS20 encodes a protein , RokLS20 , that suppresses the development of natural competence of B . subtilis [28] . Recently , we have reported a global view of the regulatory circuitry of the pLS20 conjugation genes . A conjugation operon encompassing more than 40 genes is located next to a divergently oriented single gene , rcoLS20 , which encodes the master regulator of conjugation responsible for keeping conjugation in the default “OFF” state . Activation of conjugation requires an anti-repressor , RapLS20 , that belongs to the family of Rap proteins . Inactivation of the rapLS20 gene on pLS20 severely compromises conjugation , and conjugation was enhanced when rapLS20 was expressed from an ectopic locus . The activity of RapLS20 , in turn , is regulated by a signaling peptide , Phr*LS20 . The small phrLS20 gene , located immediately downstream of rapLS20 , encodes a pre-protein . After being secreted , PhrLS20 can be processed by a second proteolytic cleavage , resulting in generation of the functional pentapeptide , Phr*LS20 , corresponding to the five C-terminal residues of PhrLS20 . When ( re ) imported , this peptide inactivates RapLS20 . Therefore , activation of conjugation is ultimately regulated by the Phr*LS20 signaling peptide . The Phr*LS20 concentration will be relatively high or low when donor cells are predominantly surrounded by donor or recipient cells , respectively . Hence , conjugation will become activated particularly under conditions in which recipient cells are potentially present . In addition , Phr*LS20 has a crucial role in returning conjugation to the default “OFF” state [29] . Despite identification of the players involved in regulation of the conjugation genes , our knowledge on regulation of the genetic switch responsible for activating conjugation is still very limited . Using a combination of various in vitro and in vivo approaches , we show that the genetic switch controlling pLS20 conjugation involves at least three layers of regulation . Together , they tightly repress the main conjugation promoter under conditions that do not favor conjugation , while maintaining the ability to accurately switch on the conjugation genes when appropriate conditions occur . The three layers involve coinciding or overlapping divergent promoters of different strengths , autoregulated expression of RcoLS20 , which turns out to be a tri-functional transcriptional regulator , and formation of RcoLS20-mediated DNA looping . The sophisticated regulatory mechanism that combines three layers of control into a single switch is novel for plasmids of Gram-positive bacteria . The implications of the uncovered regulation mechanisms for conjugation are discussed in the context of regulatory systems present on other HGT elements and with other regulatory systems involving DNA looping .
Conjugation is a complex and energy consuming process , involving the generation and transfer of ssDNA , synthesis and assembly of a sophisticated type IV secretion system , and establishment of specific contacts with the recipient cell . Hence , the process of conjugation and expression of the genes involved are strictly controlled . Analysis of the regulation of conjugation genes present on ICEs in bacteria and those on plasmids of Gram-negative bacteria indeed indicates that this is the case [for review see] , [ 5 , 7] . In our previous studies , we have sequenced and annotated plasmid pLS20cat of the Gram-positive bacterium B . subtilis and identified a large conjugation operon . We have also identified rcoLS20 as the gene encoding the master regulator of conjugation , RapLS20 as the anti-repressor required to activate the conjugation genes , and we showed that the activity of RapLS20 is in turn regulated by the signaling peptide Phr*LS20 . In this study , we analyzed the underlying molecular mechanism of how the pLS20 conjugation genes are regulated . The results obtained provide compelling evidence that the conjugation genes of pLS20 are controlled by a complex genetic switch , which is composed of at least three intertwined layers . A scheme of the three layers is shown in Figure 9 . One of the levels results from the relative positioning of the main conjugation promoter , Pc , and the divergently oriented promoter Pr , driving expression of the rcoLS20 gene ( Fig . 9A ) . The presence of divergently oriented promoters is a common form of gene organization in bacteria , and the ( likely ) role of this organization in transcriptional regulation has long been recognized [34] . Nevertheless , direct proof for and detailed analysis of the implications on transcriptional regulation are restricted to only a minor fraction of the divergently oriented transcriptional units detected . Here , we identified the conjugation promoter Pc and showed that it is a relatively strong promoter , which is repressed by the master regulator of conjugation RcoLS20 . Importantly , the position of promoter Pc coincides , or at least partially overlaps , with the divergently oriented weak Pr promoter . It has been demonstrated that an RNA polymerase can bind only to one of two overlapping promoters [35] , [36] . Thus , in the special configuration of overlapping promoters the RNA polymerase may itself act as a transcriptional regulator . Recently , Bendtsen et al . [37] described theoretical scenarios backed up by experimental data that overlapping promoters indeed can result in a transcriptional switch , provided that they have different activities in the absence of the regulatory protein , combined with a regulator that has a strong differential effect on the regulation of both promoters . This is exactly the case for the Pc/Pr promoter pair; in the absence of the regulator promoter Pc is several hundred folds stronger than Pr , and the presence of the regulator strongly represses the Pc promoter while activating the Pr promoter . The second level of regulation contributing to the genetic switch concerns the multiple roles that RcoLS20 plays in the Pc/Pr regulation ( Fig . 9B ) . We showed that , on the one hand , RcoLS20 activates transcription of its own weak promoter , Pr , thereby generating a self-sustaining positive feedback loop . On the other hand , RcoLS20 functions simultaneously as an efficient repressor of the Pc promoter . The dual effect that RcoLS20 has on Pc and Pr maintains conjugation effectively in the “OFF” state . We also showed that the level of rcoLS20 induction from an inducible promoter required for efficient repression of the Pc promoter was about ten-fold lower than that required for maximum auto-activation of the Pr promoter . These differential effects of RcoLS20 on repressing and activating the Pc/Pr promoters will also contribute towards maintaining conjugation stably in the “OFF” state under conditions when conjugation should not be activated . Interestingly , we found that at elevated concentrations RcoLS0 inhibits its own transcription . This negative autoregulation probably functions to keep RcoLS20 within a low concentration range in order to respond accurately to the anti-repressor RapLS20 to activate the conjugation genes . The triple effects RcoLS20 has on the regulation of the Pc/Pr promoters will also play an important role when RapLS20 induces the system to switch to the “ON” state . In addition to relieving repression of the strong conjugation Pc promoter , this will simultaneously annihilate autostimulation of the Pr promoter , preventing further synthesis of RcoLS20 , which in turn will contribute in pushing and maintaining conjugation in the “ON” state . A third level contributing to the genetic switch to activate the conjugation genes involves the DNA looping mediated by simultaneous binding of RcoLS20 to operators OI and OII ( Fig . 9C ) . DNA looping mediated by a transcriptional regulator has been reported for several other regulatory systems in prokaryotes and their analyses have revealed that several features are conserved and necessary for DNA looping to occur [for review see , 38] . Our results showed that the properties of RcoLS20 and the DNA in the Pc/Pr region comply with the necessary features for RcoLS20-mediated loop formation . First , using different techniques , we show that RcoLS20 , -predicted to contain a helix-turn-helix DNA binding motif in its N-terminal region [29]- , is a DNA binding protein and that it binds specifically to two operators , OI and OII . Second , operator OI , which is located more than 85 bp away from promoters Pc and Pr , is required for efficient regulation of both promoters . Third , RcoLS20 binds cooperatively to both operators . Fourth , dephasing the positions of the two operators by inserting 5 bp in the spacer region destroys proper regulation of the conjugation genes . And fifth , we showed that RcoLS20 forms tetramers in solution . This will create a unit containing multiple DNA binding motifs , facilitating cooperative binding to multiple sites within the two operators . The DNA loop in the Pc/Pr region of pLS20 is characterized by a small spacer region that separates RcoLS20 operators OI and OII . The spacer length can be used to classify DNA loops into two categories: short or energetic loops , and long or entropic ones . Due to intrinsic stiffness and torsional rigidity of the DNA , loop formation is normally unfavorable for those with spacer lengths shorter than the DNA persistence length ( approximately 150 bp ) , because the curvature energy required for forming such small loops becomes too great . For such short loops to occur specific features like intrinsic static bending or binding of an additional protein inducing bending are required . In the case of pLS20 , in which the operators OI and OII are separated by only 75 bp , we show that the spacer region contains a static bent . The first experimental demonstration that a DNA loop can play a crucial role in transcriptional regulation was reported for the E . coli ara operon in 1984 [39] . Since then , some other operons have been shown to be also regulated by transcriptional regulator-mediated DNA looping [for reviews see] , [ 38] , [40]–[43] , though the actual number of transcriptional systems for which DNA looping has been conclusively demonstrated is remarkably low . In the case of plasmids , reports demonstrating DNA looping systems are limited to only few cases . One of these includes regulation of initiation of DNA replication at the beta origin of the E . coli R6K plasmid [44]; and in the case of Enterococcus faecalis plasmid pCF10 it has been proposed that regulation of its conjugation system involves DNA looping mediated by the pheromone-responsive transcriptional regulator PrgX [for review see , 45] . Bio-informatic analyses suggest that DNA looping mediated regulation of transcription is likely to be more common than the few cases for which this has been demonstrated so far . For instance , Cournac and Plumbridge [38] have screened the E . coli genome for the presence of putative “simple DNA looping systems” in which looping would involve a single regulator ( i . e . , this analysis included only transcriptional regulators for which the operator sequence is known , and did not take into account the putative loops that would involve heterologous proteins and/or global transcriptional regulators ) . Under these restrictive settings , this survey identified 48 genes/operons in which DNA looping mediated regulation is likely to play a role . Interestingly , fourteen of them involve divergently oriented promoters . In the context of our studies , it is worth mentioning the regulation of the conjugation genes located on the integrative and conjugative element ICEBs1 that is present in several B . subtilis strains . The gene encoding the transcriptional regulator ImmR , and the excision and conjugation genes are expressed from two divergently oriented promoters that are separated by ∼130 bp . At low concentrations , the ImmR protein can bind to six regions , three being proximal to each promoter . It has been suggested that repression of the immR promoter might involve cooperative interactions between ImmR molecules bound to binding sites proximal to both promoters , i . e . DNA looping [46] . Based on the distribution of the operator sites , DNA looping could also be involved in the transcriptional regulation of the Gram-negative plasmids Ti or IncP-plasmids , where divergent promoters have been shown to be involved in controlling both the replication and transfer functions [47] , [48] . What are the benefits of DNA looping in general and for the regulation of the conjugation genes of pLS20 in particular ? A major consequence of DNA looping is that it results in a high local concentration of the transcriptional regulator at the right place , which would increase its specificity and affinity [for recent review see , 49] . Often , -and RcoLS20 is not an exception- , transcriptional regulators are produced in limited amounts per cell . Low numbers of regulators enhance the possibility of transcriptional fluctuations between individual cells within a population . In addition , the intrinsic stochasticity of transcription , -also referred to as noise- , affects the temporal effectiveness of transcriptional regulation; again this is especially prominent when the number of regulatory proteins involved is low . Recent evidences indicate that DNA looping contributes importantly to controlling temporal transcriptional noise , as well as dampening transcriptional fluctuations between cells within a population [50] , [51] . Thus , DNA looping contributes to the tight regulation of promoters especially when levels of transcriptional regulators are low by diminishing stochastic fluctuations in transcription . For some differentiation processes , cell-to-cell or stochastic variability in levels of transcriptional regulators form the basis for activation of these processes , resulting in different behavior of genetically identical cells within a population [52]–[54] . Examples of these processes are the formation of persister cells , development of natural genetic competence , spore formation and swimming/chaining . It is believed that such a bet-hedging strategy is beneficial for the fitness of the species because there will always be some cells that are prepared to cope with a deteriorating environmental condition that may arise in the near future . However , for other processes , there may not be such an advantage and it would then be important to tightly repress the process at times when conditions for that process are not apt . Conjugation probably is such a process because there is no benefit in activating the conjugation genes when there is no recipient present to receive the plasmid . The fact that the efficiency of pLS20 transfer during growth conditions antithetic to conjugation is below the detection limit ( at least six orders of magnitude lower than those observed during optimal conjugation conditions ) strongly indicates that conjugation genes are tightly repressed under such conditions . However , the tight repression of conjugation should not compromise the ability of rapidly switching to high expression of the conjugation genes when appropriate conditions occur . In pLS20 this is achieved by the constellation of DNA looping combined with autoregulated expression of RcoLS20 and overlapping divergent promoters of different strength . A well-studied genetic switch involving DNA looping is the one that governs the switch from the lysogenic to the lytic state of the Escherichia coli phage λ [for review see] , [ 55 , 56] . In the lysogenic or prophage state , phage λ replicates passively with the host while the lytic genes are repressed . This prophage state is extremely stable and can be maintained for many generations . Upon induction of the SOS response , however , a switch is made to the lytic cycle resulting in excision of the phage genome , followed by its amplification and eventually lysis of the cell and release of phage progeny . The early lytic phage λ genes are located in two divergently oriented operons , which are controlled by the lytic promoters PR and PL . A third operon , which encodes amongst others the CI transcriptional regulator , is located in between the two early lytic operons such that the promoter of gene cI , PRM , flanks the divergently oriented PR promoter driving expression of one of the two early operons . In several aspects , functional analogies exist between CI and RcoLS20 although they share only 16% of identity at their primary protein sequence level . Both RcoLS20 and CI stimulate and repress their own promoter at low and high concentrations , respectively , resulting in a self-sustaining positive feedback loop while keeping the transcriptional regulator in a low concentration range . Above , arguments have been given that for pLS20 this situation , together with the effects of the DNA loop , is important for the tight repression of the Pc promoter during conditions in which conjugation is not favourable , while maintaining the sensitivity to be able to respond rapidly to switch on the conjugation genes when appropriate conditions occur . The transcriptional regulation of λ appears to serve a similar purpose . Thus , on the one hand the lytic genes are tightly repressed since spontaneous switching to the lytic cycle occurs less than once every 108 generations [57] . On the other hand , mutations that specifically eliminate the negative autoregulation of cI expression impair prophage induction [58] , [59] . Another analogy between the pLS20 and λ systems is that both the regulators RcoLS20 and CI , can form higher order oligomers , permitting them to bind cooperatively to multiple sites distributed in two operators , effectively resulting in DNA looping which plays an important role in the genetic regulation of the conjugation and the lytic operon , respectively . Taking the analogy further , it is interesting to note that these regulatory systems both control a process of horizontal gene transfer . However , there are also several differences between the two systems . For instance , whereas regulation of pLS20 conjugation genes involves a short loop of 75 bp , regulation of the λ lytic genes involves a long loop of 2 . 3 kb . A second difference is that CI protein forms dimers in solution . A pair of CI dimers tetramerizes when binding to the binding sites in one operator and another dimer pair does the same when binding to the other operator . Upon DNA looping , interaction between the two tetramers constitutes a functional octamer . In addition , when a loop is formed another pair of dimers may bind to additional binding sites present in both operators , and this additional bridge is responsible for repressing PRM promoter . At present , we do not have such detailed insights in transcriptional regulations at the molecular level for RcoLS20 . However , instead of dimers , RcoLS20 forms tetramers in solution , which probably means that the molecular mechanism by which the pLS20 promoters Pr and Pc are regulated is distinct from the way CI regulates λ promoters PR and PRM . Another argument supporting this assumption is the different configuration of the divergent promoters and the binding sites for the regulator protein . In pLS20 , the position of promoters Pc/Pr overlaps and the RcoLS20 binding sites in OII overlap and flank these core promoters . In λ the binding sites for CI regulator in one operator overlap the PR promoter and are located upstream of the PRM core promoter sequences . Finally , a major difference between the DNA looping involved systems of pLS20 and λ is how the switches are induced . In λ , the switch is induced by an SOS response which results in RecA-mediated CI autocleavage . In the case of pLS20 , the switch is dictated ultimately by intercellular quorum sensing signaling involving the signaling peptide Phr*LS20 that regulates the activity of RapLS20 , the anti-repressor of RcoLS20 [29] . This quorum sensing system will lead to activation of the conjugation genes when donor cells are surrounded by recipient cells . However , high levels of Phr*LS20 will build up when the majority of the cells that surround a donor cell already contain pLS20 , and this will inactivate RapLS20 and hence block activation of the conjugation genes . Besides those described here , it is possible that the pLS20 conjugation genes are regulated by additional mechanism ( s ) . For example , the conspicuously long 5′ untranslated region upstream of gene 28 is predicted to form complex secondary structures , which might modulate expression of the downstream genes in a variety of scenarios . Currently , a study to elucidate a possible role of this long 5′ untranslated region is carried out in our laboratory . In summary , in this work we have provided evidence that regulation of the conjugation genes present on pLS20 is based on a unique genetic switch that combines at least three levels of control . These include ( i ) overlapping divergent promoters of different strengths , ( ii ) auto-stimulation and repression of the weak Pr promoter by the transcriptional regulator at low and elevated concentrations , respectively , combined with simultaneous repression of the divergent strong conjugation promoter , and ( iii ) DNA looping mediated by binding of RcoLS20 regulator to two operators separated by a short loop . Most likely , the combination of these different layers causes tight repression of the main conjugation promoter Pc when conditions for conjugation are not optimal , while allowing the system to switch rapidly to high expression of the conjugation genes when appropriate conditions occur .
Bacterial strains were grown in LB liquid medium or on 1 . 5% LB agar plates [60] . When appropriate , the following antibiotics were added to media or plates: ampicillin ( 100 µg/ml ) , erythromycin ( 1 and 150 µg/ml in B . subtilis and E . coli , respectively ) , chloramphenicol ( 5 µg/ml ) , spectinomycin ( 100 µg/ml ) , and kanamycin ( 10 µg/ml ) . Table S1 lists the B . subtilis strains used . All of them are isogenic with B . subtilis strain 168 . Plasmids and oligonucleotides used are listed in Table S2 and S3 , respectively . All oligos were purchased from Isogen Life Science , The Netherlands . E . coli cells were transformed using standardized methods as described in Singh et al [61] . For standard B . subtilis transformations , competent cells were prepared as described by Bron [62] . Transformants were selected on LB agar plates with appropriate antibiotics . Standard molecular methods were used to manipulate DNA [60] . Sequence analysis was used to verify the correctness of all constructs . The same strategy was used to construct B . subtilis strains containing a copy of lacZ fused to the entire or part of the rcoLS20-gene 28 intergenic DNA region . First , the region of DNA to be cloned was amplified using appropriate primers ( see Table S3 ) , purified , and digested with the appropriate restriction enzymes . Next , the fragment was used to prepare a ligation mixture together with the integration vector pDG1663 digested with the same enzymes . The ligation mixture was transformed into E . coli XL1-blue cells . The plasmid content of several ampicillin resistant transformants was checked and clones containing the insert with appropriate size and orientation were subjected to DNA sequencing to verify the absence of mutations . The names of the pDG1663 derivatives and their characteristics are listed in Table S2 . Plasmid DNA of each pDG1663 derivative was used to transform competent B . subtilis 168 cells . Transformants were initially selected for resistance to erythromycin . Next , double cross-over events were distinguished from single cross-over events by selecting transformants sensitive to spectinomycin . The resulting B . subtilis strains containing a single copy of lacZ preceded by different regions of the rcoLS20-gene 28 region at the thrC locus of the B . subtilis chromosome are listed in Table S1 . Next , plasmid pLS20cat was introduced into the different lacZ fusion strains by conjugation . B . subtilis strain PKS9 contains a single copy of the rcoLS20 gene under the control of the IPTG-inducible Pspank promoter at its amyE locus and this cassette is linked to the spectinomycin gene . Chromosomal DNA of strain PKS9 was used to transform competent cells of the various lacZ fusion strains in order to construct derivatives of the lacZ fusion strains containing the Pspank-rcoLS20 cassette . The following strategy was used to construct a translational fusion of rcoLS20 with his ( 6 ) . The rcoLS20 gene was amplified from pLS20cat by PCR using primers oPKS14N and oPKS8 . The purified PCR product was digested with NcoI and SalI and cloned into the vector pET28b+ digested with the same restriction enzymes to produce plasmid pRcoLS20-His . B . subtilis strain GR90 contains the rcoLS20-his ( 6 ) under the control of the Pspank promoter at the amyE locus . To construct this strain rcoLS20-his ( 6 ) was amplified from pRcoLS20-His by PCR using primers oGR3 and oGR4 . The PCR product was digested with NheI and SphI and cloned into the vector pDR110 digested with the same enzymes to generate pPspankrcoLS20-His . This plasmid was used to transform competent B . subtilis cells selecting for spectinomycin resistance . Double cross-over events were selected by loss of amylase gene . β-galactosidase activities were determined as described previously [63] . Overnight grown cultures were diluted 100 times into fresh prewarmed medium and samples were taken every 45 min . Conjugation was carried out in liquid medium as described previously [29] . The effect of ectopic expression on conjugation of a gene controlled by the IPTG-inducible Pspank promoter was studied as follows . Overnight cultures were diluted in prewarmd LB supplemented with IPTG at the indicated concentrations to an OD600 of ∼0 . 05 . Next , samples were taken at regular intervals to determine OD600 and were subjected to matings with proper recipient cells . Preparation of total RNA samples , RNA sequencing and Bioinformatic analysis of RNAseq data was done as described previously [29] . E . coli BL21 ( DE3 ) cells carrying plasmid prcoLS20-His were used to inoculate 1 litre of fresh LB medium supplemented with 30 mg/ml kanamycin and grown at 37°C with shaking . At an OD600 of 0 . 4 , expression of rcoLS20-his ( 6 ) was induced by adding IPTG to a final concentration of 1 mM and growth was continued for 2 h . Cells were further processed as described previously [28] . Purified protein ( >95% pure ) was dialysed against buffer B ( 20 mM Tris-HCl pH 8 . 0 , 1 mM EDTA , 250 mM NaCl , 10 mM MgCl2 , 7 mM β-mercaptoethanol , 50% v/v glycerol ) and stored in aliquots at −80°C . Bradford assay was used to determine the protein concentrations . In essence , the gel retardation assays were carried out as described before [28] . Thus , different fragments of intergenic regions between gene 28 and rcoLS20 were amplified by PCR using pLS20cat as template . The resulting PCR fragments were purified and equal concentrations ( 300 nM ) were incubated on ice in binding buffer [20 mM Tris HCl pH 8 , 1 mM EDTA , 5 mM MgCl2 , 0 . 5 mM DTT , 100 mM KCl , 10% ( v/v ) glycerol , 0 . 05 mg ml−1 BSA] without and with increasing amounts of purified RcoLS20His ( 6 ) in a total volume of 16 µl . After careful mixing , samples were incubated for 20 min at 30°C , placed back on ice for 10 min , then loaded onto 2% agarose gel in 0 . 5XTBE . Electrophoresis was carried out in 0 . 5X TBE at 50 V at 4°C . Finally , the gel was stained with ethidium bromide , destained in 0 . 5XTBE and photographed with UV illumination . Determination of the transcription start sites by primer extension was performed essentially as described [64] . In brief , total RNA ( 30 µg ) was mixed with 4 pmol of end-labeled oligonucleotide that served as primer; the mixture was heated at 70°C for 5 min and allowed to anneal for 5 min at 23°C . The annealed RNA was ethanol precipitated , resuspended and primer extension was performed with 30 U of AMV reverse transcriptase ( Promega ) at 42°C , as recommended by the supplier . The extended cDNA products were analysed by electrophoresis on a denaturing 6% urea-polyacrylamide gel , in parallel with a DNA sequence ladder performed by chemical sequencing [65] of a DNA fragment encompassing the mapped promoters ( see below ) . The primer used to map promoter Pc was 5′-ttctagttctttttacac , while that used for promoter Pr was 5′-tctctattgcccacttat . Oligonucleotides were end-labeled with [γ-32P]-ATP and T4 polynucleotide kinase as recommended by the supplier ( New England Biolabs ) . The 186 bp DNA fragment that served as sequence ladder was PCR amplified with primers 5′-acggtctagcgcttacaat and 5′-ttctagttctttttacac , the last one labeled at its 5′ end . DNaseI footprinting assay was carried out as described [66] . The Pc/pr promoter encompassing region was amplified by PCR using primers p28_Δ16 and Prom28UpBam , and pLS20cat as template . One of the ends was radio-labeled by digesting the fragment with BamHI and subsequently filling in the end with exo− Klenow fragment in the presence of [α-32P]-ATP . Presence of conserved motifs was searched by using motif-identification programs MEME [30] and BIOPROSPECTOR [31] . Prediction of the static bending properties of DNA sequences was carried out by calculating the global 3D structure according to the dinucleotide wedge model [67] . All graphics work was done by using Adobe Photoshop CS2 and adobe illustrator . Graphs were plotted using Excel program . Sedimentation velocity assay . Samples in 20 mM Tris-HCl , 250 mM NaCl , 10 mM MgCl2 , 1 mM EDTA and 100 mM glycerol , pH 7 . 4 , were loaded ( 320 µL ) into analytical ultracentrifugation cells . The experiments were carried out at 43–48 krpm in an XL-I analytical ultracentrifuge ( Beckman-Coulter Inc . ) equipped with UV-VIS absorbance and Raleigh interference detection systems . Sedimentation profiles were recorded at 280 nm . Sedimentation coefficient distributions were calculated by least-squares boundary modelling of sedimentation velocity data using the continuous distribution c ( s ) Lamm equation model as implemented by SEDFIT 14 . 1 [68] . Experimental s values were corrected to standard conditions ( water , 20°C , and infinite dilution ) using the program SEDNTERP [69] to get the corresponding standard s values ( s20 , w ) . Sedimentation equilibrium assay . Using the same experimental conditions as in the SV experiments , short columns ( 90 µL ) SE experiments were carried out at speeds ranging from 7 , 000 to 10 , 000 rpm and at 280 nm . After the last equilibrium scan , a high-speed centrifugation run ( 48 , 000 rpm ) was done to estimate the corresponding baseline offsets . Weight-average buoyant molecular weights of protein were determined by fitting a single species model to the experimental data using the HeteroAnalysis program [70] , and corrected for solvent composition and temperature with the program SEDNTERP [69] .
|
Plasmids are extrachromosomal , autonomously replicating units that are harbored by many bacteria . Many plasmids encode transfer function allowing them to be transferred into plasmid-free bacteria by a process named conjugation . Since many of them also carry antibiotic resistance genes , plasmid-mediated conjugation is a major mechanism in the dissemination of antibiotic resistance . In depth knowledge on the regulation of conjugation genes is a prerequisite to design measures interfering with the spread of antibiotic resistance . pLS20 is a conjugative plasmid of the soil bacterium Bacillus subtilis , which is also a gut commensal in animals and humans . Here we describe in detail the molecular mechanism by which the key transcriptional regulator tightly represses the conjugation genes during conditions unfavorable to conjugation without compromising the ability to switch on accurately the conjugation genes when appropriate . We found that conjugation is subject to the control of a unique genetic switch where at least three levels of regulation are integrated . The first level involves overlapping divergent promoters of different strengths . The second layer involves a triple function of the transcriptional regulator . And the third level of regulation concerns formation of a DNA loop mediated by the transcriptional regulator .
|
[
"Abstract",
"Introduction",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"microbiology",
"evolutionary",
"biology",
"molecular",
"biology"
] |
2014
|
A Complex Genetic Switch Involving Overlapping Divergent Promoters and DNA Looping Regulates Expression of Conjugation Genes of a Gram-positive Plasmid
|
Many persistent transmitted plant viruses , including rice stripe virus ( RSV ) , cause serious damage to crop production worldwide . Although many reports have indicated that a successful insect-mediated virus transmission depends on a proper interaction between the virus and its insect vector , the mechanism ( s ) controlling this interaction remained poorly understood . In this study , we used RSV and its small brown planthopper ( SBPH ) vector as a working model to elucidate the molecular mechanisms underlying the entrance of RSV virions into SBPH midgut cells for virus circulative and propagative transmission . We have determined that this non-enveloped tenuivirus uses its non-structural glycoprotein NSvc2 as a helper component to overcome the midgut barrier ( s ) for RSV replication and transmission . In the absence of this glycoprotein , purified RSV virions were unable to enter SBPH midgut cells . In the RSV-infected cells , this glycoprotein was processed into two mature proteins: an amino-terminal protein ( NSvc2-N ) and a carboxyl-terminal protein ( NSvc2-C ) . Both NSvc2-N and NSvc2-C interact with RSV virions . Our results showed that the NSvc2-N could bind directly to the surface of midgut lumen via its N-glycosylation sites . Upon recognition , the midgut cells underwent endocytosis followed by compartmentalization of RSV virions and NSvc2 into early and then late endosomes . The NSvc2-C triggered cell membrane fusion via its highly conserved fusion loop motifs under the acidic condition inside the late endosomes , leading to the release of RSV virions from endosomes into cytosol . In summary , our results showed for the first time that a rice tenuivirus utilized its glycoprotein NSvc2 as a helper component to ensure a proper interaction between its virions and SBPH midgut cells for its circulative and propagative transmission .
Arthropod insects play critical roles in epidemics of numerous animal and plant viruses [1–3] . Based on the mode of transmission , plant viruses can be classified into non-persistent , semi-persistent or persistent transmitted viruses [4–6] . For non-persistent or semi-persistent transmissions , plant viruses are retained inside insect stylets from a few minutes to several hours or on insect foregut surface for a few hours to several days . Upon probing or feeding on a host plant , viruses are quickly injected into plant cells , together with insect saliva [7–9] . The persistent transmitted plant viruses ( non-propagative or propagative ) can enter insect vector bodies , and then circulate and/or replicate inside the vectors for several days to weeks [10] . These persistent transmitted viruses need to pass insect midgut barrier ( s ) , dissemination barrier ( s ) , and then salivary gland barrier ( s ) prior to be transmitted to new host plants [6 , 11 , 12] . Insect midgut is often considered to be one of the major barriers for successful persistent virus transmissions . During the process of passing through midgut barrier ( s ) , proper interactions between viruses and vectors are needed . To date , the mechanism ( s ) controlling the interactions between viruses and their insect vector midgut barrier ( s ) are poorly understood . Helper component-mediated mechanisms have been reported for the non-persistent , semi-persistent and persistent-nonpropagative transmitted plant viruses , respectively [4 , 13–15] . For example , virions of non-persistent or semi-persistent transmitted plant viruses were reported to interact with the cuticular proteins inside the insect mouthpart or foregut [16 , 17] , and these virus–insect interactions required virally encoded non-structural helper factors as molecular bridges [13 , 14] . Viruses in the genus Potyvirus are known to encode a helper component proteinase ( HC-Pro ) that can act as a molecular bridge for the interaction between potyvirus virions and its aphid vectors [18–20] . Members in the genus Caulimovirus encode a different helper factor that can help virions to retain on insect maxillary stylets [21–23] . Virions of multiple persistent ( including propagative and non-propagative ) transmitted plant viruses ( e . g . , luteovirus [24 , 25] , geminivirus [26 , 27] , reovirus [28 , 29] , tospovirus [30 , 31] , and plant rhabdovirus [32 , 33] ) were reported to bind directly to insect midgut cells , whereas these bindings depended on virions surface-exposed proteins . Faba bean necrotic yellows virus , a persistent-nonpropagative nanovirus , was found to require a helper factor for transmission by its aphid vector . To date , however , no persistent-propagative transmitted plant viruses were reported to rely on virally encoded helper proteins for their transmission . Rice stripe virus ( RSV ) is transmitted by SBPH in a circulative and propagative manner , and often causes severe losses to rice production in China and many other countries in Asia [34 , 35] . The genome sequence of plant-infecting tenuivirus is similar to the members of animal-infecting Phlebovirus in the order of Bunyavirales . Most members in the order Bunyavirales are known to produce membrane-enveloped spherical virions with two surface-exposed glycoproteins , and these glycoproteins are important for virus entrance into host cells or for vector transmission [31 , 36 , 37] . Virions of tenuiviruses are filamentous and do not have envelope membranes [38–40] . RSV also encodes a glycoprotein NSvc2 ( 92 kDa ) , which is further processed into an amino-terminal part protein known as NSvc2-N ( 40 kDa ) and a carboxyl-terminal part protein known as NSvc2-C ( 50 kDa ) [41 , 42] . However , this glycoprotein is not present in the purified RSV virions [43 , 44] . Based on the published reports , we hypothesized that RSV must use a different strategy to overcome the midgut barrier ( s ) for its insect transmission . To validate this hypothesis , we conducted multiple experiments on the interaction between RSV and SBPH during virus entrance into insect vector midgut . We have now determined that this virus uses a viral glycoprotein NSvc2 as a helper component to overcome SBPH midgut barrier ( s ) for its persistent-propagative transmission . We have also determined that in the absence of NSvc2 , RSV virions were unable to enter SBPH midgut cells . Our results further demonstrated that this glycoprotein acted as a critical helper component to ensure the proper interaction between RSV virions and SBPH midgut cells . Both NSvc2-N and NSvc2-C interacted with RSV virions and NSvc2-N bound directly to the midgut barrier ( s ) . Upon successful interaction , the midgut cells underwent endocytosis followed by compartmentalization of RSV virions , NSvc2-N and NSvc2-C complexes ( referred to RSV virions:NSvc2-N:NSvc2-C complex thereafter ) inside the early and then late endosomes . NSvc2-C triggered membrane fusion under the acidic condition inside the late endosomes to release RSV virion:NSvc2-N complexes into the cytosol . These new findings expanded our knowledge on the interactions between virions and their insect vectors during plant virus persistent transmissions .
To examine whether NSvc2 plays important role ( s ) in RSV circulative-propagative transmission , we first conducted a time course study on the localizations of NSvc2 and RSV virions inside the midgut of SBPH during RSV acquisition . SBPHs were fed on RSV-infected rice seedlings and then collected at 4 , 8 , 16 and 24 h post feeding ( 30 SBPHs per time point ) , respectively . The collected insects were dissected and analyzed for the presence of NSvc2 and RSV virions by a double-immunolabeling using antibodies against RSV NSvc2-N or virion-surface nucleocapsid protein ( NP ) . As shown in Fig 1A , RSV virions ( green ) had accumulated inside the midgut lumen at 4 h post feeding on the RSV-infected rice seedlings . In the same tissues , NSvc2 was also detected ( red ) and some of the NSvc2 signals were co-localized with RSV virions on the actin antibody-labelled intestinal microvillus ( blue ) ( Fig 1A and S1A Fig ) , which are cellular membrane protrusions composed of dense bundles of cross-linked actin filaments at the surface of midgut epithelial cells [45] . The overlapped coefficient ( OC ) value for the red and green labeling signals was 0 . 76 ± 0 . 03 ( Fig 1E ) , indicating that NSvc2 and RSV virions were localized close to each other . At 8 h post feeding , NSvc2 was found to co-localize with RSV virions in various sized vesicle-like structures inside the epithelial cells ( Fig 1B and S1A Fig ) . Analysis of the OC value showed again that NSvc2 and RSV virions were localized close to each other ( Fig 1F ) . At 16 h post feeding , RSV virions were detected together with NSvc2 in the cytosol of the midgut epithelial cells , with an OC value of 0 . 73 ± 0 . 06 ( Fig 1C and 1G and S1A Fig ) . Even at 24 h post feeding , NSvc2 was still co-localized with RSV virions , with an OC value of 0 . 76 ± 0 . 04 ( Fig 1D and 1H and S1A Fig ) . The SBPHs fed on the healthy rice seedlings did not give positive signals for NSvc2 ( red ) and RSV virions ( green ) ( S2A and S2B Fig ) . These data indicated that the RSV-encoded NSvc2 was associated with RSV virions inside SBPH midgut during insect feeding on RSV-infected rice plants . RSV virions were purified from RSV-infected rice seedlings through ultracentrifugation using a 20% glycerol cushion . After ultracentrifugation , four different supernatant fractions starting from the top ( Sup1 to Sup4 ) , four 20% glycerol phase fractions starting from the top ( Gly1 to Gly4 ) , and the pellet ( Pel ) were collected and analyzed individually by immunoblotting assays using an antibody against RSV NP ( 35 kDa ) or NSvc2-N ( Fig 2A and 2B ) . Results showed that the resuspended pellet sample contained RSV virions , and the four supernatant fractions ( Sup1 to Sup4 ) contained NSvc2 protein only . In contrast , the four glycerol phase fractions ( Gly1 to Gly4 ) contained both RSV virions and NSvc2 ( Fig 2B ) . Transmission Electron Microscopy showed that numerous filamentous RSV virions were present in the resuspended pellet sample ( Fig 3A ) . Immunogold labeling using a gold-labeled NP specific antibody showed that the immunogold particles were attached to the purified RSV virions ( Fig 3B ) . When a NSvc2 specific antibody was used in the immunogold labeling assay , no immunogold particle was observed together with the purified RSV virions ( Fig 3D to 3F ) , confirming that the purified RSV virions do not contain NSvc2 . To further verify the above findings , SBPHs were allowed to feed on a mixture of sucrose and the combined supernatant fractions , sucrose and the combined glycerol fractions , sucrose and the resuspended pellet sample , sucrose alone , or the combined supernatant fractions and the resuspended pellet samples through two layers of stretched parafilm for 24 h . As shown in Fig 2D ( row 2 ) , NSvc2 ( red ) and RSV virions ( green ) were detected together inside the epithelial cells of SBPHs fed on the mixture containing sucrose and the combined glycerol fractions . NSvc2 was , but not RSV virions , detected inside the midgut lumen of the SBPHs fed on the mixture containing sucrose and the combined supernatant fractions ( Fig 2D , row 1 ) . RSV virions were detected inside the midgut lumen but not inside the epithelial cells of the SBPHs fed on the mixture containing sucrose and the resuspended pellet samples ( Fig 2D , row 3 ) . When insects fed on the mixture containing sucrose , the combined supernatant fractions and the resuspended pellet samples , however , both NSvc2 and RSV virions were detected inside the epithelial cells ( OC value = 0 . 96 ± 0 . 04; Fig 2D , row 4 ) , similar to that found for the SBPHs fed on the mixture containing sucrose and the combined glycerol phase fractions ( OC value = 0 . 97 ± 0 . 03 ) . No NSvc2 or RSV virions was detected inside the midgut of the SBPHs fed on the mixture containing sucrose and the combined glycerol phase fractions from the healthy rice seedlings ( S2C Fig ) . To further investigate the roles of NSvc2 in RSV acquisition and transmission by SBPH , SBPHs were fed with a mixture containing sucrose and the combined supernatant fractions , a mixture containing sucrose and the combined glycerol phase fractions , a mixture containing sucrose and the resuspended pellet sample , or a mixture containing sucrose , the mixed supernatant fractions and the resuspended pellet sample for 48 h and then on rice seedlings . The rate of SBPH acquired RSV and the rate of SBPH transmitted RSV to rice seedlings were determined by immuno-dot blot assays . The results showed that after feeding on the mixture containing sucrose , the combined supernatant fractions and the resuspended pellet samples , RSV virions were successfully transmitted to rice seedlings ( Fig 2C and S1 Table ) . No RSV acquisition and transmission were observed for the SBPHs fed on sucrose only or on the mixture containing sucrose and the resuspended pellet sample . We then performed yeast two-hybrid and co-immunoprecipitation assays . Our results showed that both NSvc2-N and NSvc2-C interacted with RSV NP ( Fig 2E and 2F ) , suggesting that RSV NSvc2 played a critical role in mediating the entrance of RSV virions into SBPH midgut cells . A previous research has shown that NSvc2 can be further processed into two mature glycoproteins , namely NSvc2-N and NSvc2-C [42] . S3A Fig illustrated the predicted positions of the signal peptide ( SP ) , transmembrane ( TM ) regions and glycan sites within NSvc2 . To investigate the roles of NSvc2-N in RSV transmission , we deleted the TM region and expressed the soluble part of NSvc2-N protein ( referenced to as NSvc2-N:S thereafter , S3B Fig ) in Spodoptera frugiperda ( Sf9 ) cells using a recombinant baculovirus expression system . After purification using the Ni-NTA agarose , the expression of the recombinant NSvc2-N:S was confirmed by Western blot assays using an anti-NSvc2-N polyclonal antibody ( S3C Fig ) . To confirm that NSvc2-N:S can bind midgut epidermal microvillus , SBPHs were allowed to feed on purified NSvc2-N:S for 3 h followed by a 12 h feeding on a sucrose solution to remove unbound NSvc2-N:S . Results of immuno-labeling assays showed that NSvc2-N:S ( green ) could be readily detected in the midgut lumen near the surface of epithelial cells along the alimentary canal ( Fig 4A ) . As a negative control , SBPHs were allowed to feed on the Tomato spotted wilt virus ( TSWV ) -encoded soluble glycoprotein ( Gn:S ) , known to bind thrip midgut cells [30] . As expected , the TSWV Gn:S ( green ) was not detected inside the SBPH midgut ( Fig 4A ) . Based on the above findings , we further hypothesized that pre-acquisition of NSvc2-N:S could prevent RSV infection through blocking the RSV specific receptors on the midgut . To test this hypothesis , SBPHs were allowed to feed on the purified NSvc2-N:S for 24 h and then on the RSV-infected rice plants for 48 h . The midguts were dissected from the SBPHs and probed with the RSV NP or NSvc2-N specific antibody . Under the confocal microscope , the labeled RSV virions were found inside the midgut lumen , while the virions in the midgut epithelial cells were significantly reduced in the SBPHs pre-fed with purified NSvc2-N:S ( S4A and S4D Fig ) . In contrast , RSV virions were detected in the midgut epithelial cells of the SBPHs pre-fed with TSWV Gn:S or with sucrose alone ( S4B–S4D Fig ) . To further elucidate the roles of NSvc2-N , SBPHs pre-fed with NSvc2-N:S were allowed to feed on the RSV-infected rice plants for 48 h followed by examinations of RSV acquisition and transmission . Immuno-dot blot assays showed that pre-feeding SBPHs with NSvc2-N:S significantly reduced the rate of RSV acquisition and transmission to rice seedlings compared with that in the SBPHs pre-fed with TSWV Gn:S or with sucrose only ( Fig 4F and S2 Table ) . We then performed quantitative RT-PCR to determine the relative levels of RSV RNA in the assayed SBPHs at different time points . The results showed that RSV accumulation in the SBPHs pre-fed with NSvc2-N:S was significantly reduced compared with that in the SBPHs pre-fed with TSWV-Gn:S or with sucrose alone ( Fig 4G ) . These findings indicated that NSvc2-N:S could inhibit RSV entrance into SBPH midgut cells . Computer-assisted modeling suggested that NSvc2 might be modified through glycosylation ( S3A Fig ) . To confirm this prediction , purified NSvc2-N:S was incubated with PNGaseF ( an N-glycosidase ) to remove the N-linked glycan or with O-Glycosidases and Neuraminidase ( O-Gly + Neur ) to remove the O-linked glycan . Subsequent SDS-PAGE and immunoblotting assays showed that the purified NSvc2-N:S protein band was shifted in the gel after the PNGaseF treatment compared with the non-treated NSvc2-N:S ( Fig 4B , compare lane 1 with lane 2 ) . No clear protein band shift was observed after the treatment with O-Gly + Neur ( Fig 4B , compare lane 1 with lane 3 ) . When NSvc2-N:S was treated with PNGaseF and O-Gly + Neur , the protein band shifted as that shown by the PNGaseF-treated NSvc2-N:S ( Fig 4B , compare lane 2 with lane 4 ) , indicating that the NSvc2-N contains the N-linked glycan . We then introduced alanine-substitution mutations at the putative N-linked glycan sites of NSvc2-N:S to produce NSvc2-N:SN114A/N199A/N232A or at the putative O-linked glycan sites of NSvc2-N:S to produce NSvc2-N:SS38A/S128A/S183A . These two mutants were purified as describe above for NSvc2-N:S followed by the enzymatic deglycosylation treatments . Results showed that , without PNGaseF treatment , mutant NSvc2-N:SN114A/N199A/N232A showed a protein band shift similar to that shown by the PNGaseF-treated NSvc2-N:S ( Fig 4C , compare lane 2 with lane 3 ) . With PNGaseF treatment , the band position of NSvc2-N:SN114A/N199A/N232A mutant further shifted ( Fig 4C , compare lane 3 with lane 4 ) . In this study , no O-linked glycan modification was detected for mutant NSvc2-N:SS38A/S128A/S183A ( Fig 4D ) . These results indicated that among the three residues ( N114 , N199 and N232 in NSvc2-N:S ) , one or more residues were indeed N-glycosylated . To investigate whether the N-linked glycosylation can affect the entrance of RSV into midgut ( s ) , SBPHs were fed with the two mutant proteins , respectively , for 3 h followed by a 12 h feeding on a sucrose solution to clean insect alimentary canals . The SBPHs fed with the NSvc2-N:SN114A/N199A/N232A mutant showed almost no labeling signal at the surface of midgut microvillus ( Fig 4E , left ) . In contrast , the SBPHs fed with the NSvc2-N:SS38A/S128A/S183A mutant showed labeling signal ( Fig 4E , right ) . To determine whether the effect of NSvc2-N on RSV acquisition and transmission could be affected by the mutation caused by N-glycosylation , we pre-fed SBPHs with the NSvc2-N:SN114A/N199A/N232A mutant and then tested RSV acquisition and transmission through immuno-dot blot assays . The results showed that the SBPHs pre-fed with this mutant protein gave similar RSV acquisition and transmission rates as the SBPHs pre-fed with sucrose only ( Fig 4F and S2 Table ) . Our quantitative RT-PCR results showed that the accumulation level of RSV RNA in the SBPHs pre-fed with the NSvc2-N:SN114A/N199A/N232A mutant was similar to that in the SBPHs pre-fed with TSWV-Gn:S or with sucrose only ( Fig 4G ) . Consequently , we conclude that modification of NSvc2-N through N-glycosylation is important for RSV accumulation in SBPHs . Endocytosis is an important process during the entrance of several circulative-transmitted animal-infecting viruses into animal cells [46] . Because the above results had indicated that RSV virions and NSvc2-N accumulated together inside vesicles-like structures in SBPH midgut cells ( Fig 1B ) , we decided to investigate whether these vesicles were endosomes using early or late endosome specific markers ( e . g . , Rab5 , EEA1 and Rab7 ) as previously reported [47 , 48] . Results of this study showed that the RSV virion labeling signal was indeed co-localized with the labeling signal from the early endosome Rab5 marker or from the late endosome Rab7 marker ( Fig 5A and 5B ) . In the same study , the labeling signal from NSvc2-N or NSvc2-C was co-localized with the labeling signal from the early endosome EEA1 marker ( Fig 5C and 5D ) . To examine the potential role ( s ) of NSvc2-C , SBPHs were allowed to feed on RSV-infected rice seedlings for 4 , 8 , 16 or 24 h , and then tested by immunofluorescence labeling assays . NSvc2-C showing red labeling signal was observed together with the labeling signal from RSV virions ( green ) at the surface of microvillus ( blue ) ( S5A and S5E Fig , OC value = 0 . 86 ± 0 . 04 ) , and in the endosomal-like vesicles inside epithelial cells ( S5B and S5F Fig , OC value = 0 . 95 ± 0 . 03 ) at 4 and 8 h post feeding . At 16 and 24 h post feeding , the labeling signal from RSV virions ( green ) was observed alone in the cytosol of the epithelial cells ( S5C and S5G Fig , OC value = 0 . 18 ± 0 . 02; S5D and S5H Fig , OC value = 0 . 32 ± 0 . 06 ) , suggesting that NSvc2-C was retained inside the endosomal-like vesicles while RSV virions were released from the vesicles into the cytosol . To confirm this finding , we double-immunolabeled RSV virions and NSvc2-N or RSV virions and NSvc2-C inside the cells and monitored the changes of the localization patterns in the endosomes starting from 16 to 24 h post feeding . The results showed that both RSV virions and NSvc2-N were released from the endosomes and stayed together in the cytosol ( Fig 5E , white dashed circles; Fig 5G , arrows ) . In contrast , NSvc2-C stayed inside the endosomes and thus was not found in the cytosol of the epithelial cells ( Fig 5F , white dashed box ) . This finding indicated that RSV virions , NSvc2-N and NSvc2-C formed complexes and entered into the endosomes , and later the RSV virions:NSvc2-N complexes were released from the endosomes into the cytosol . From early endosomes to late endosomes , the pH value inside the endosome lumen shifted from 6 . 5 to 5 . 0 [46] . The acidic condition inside the late endosomes could alter the conformations of virus proteins and trigger cell membrane fusion . To determine whether NSvc2-N and/or NSvc2-C played roles in cell membrane fusion under the acidic condition , we fused a signal peptide from a baculovirus protein ( gp64 ) to NSvc2-N and NSvc2-C and expressed these two proteins individually in Sf9 cells through the recombinant baculovirus expression system ( Fig 6A ) . We first confirmed the expressions of these two proteins by Western blot assays ( S6C Fig ) . We then conducted immune-labeling assays to further confirm the above results . Under the laser scanning confocal microscope , red labeling signal representing NSvc2-N or NSvc2-C was observed in the Sf9 cell membranes ( Fig 6B ) . We then tested NSvc2-N and NSvc2-C individually to see if they could trigger cell membrane fusion under an acidic condition . The Sf9 cells were first transfected with a baculovirus expressing NSvc2-N or NSvc2-C , and then treated with a PBS solution , pH 5 . 0 , for 2 min followed by growth in a normal medium . At 4 h post acidic PBS treatment , numerous cell-cell fusions ( syncytia ) were observed for the Sf9 cells transfected with the baculovirus expressing NSvc2-C ( Fig 6E ) . In contrast , no significant cell-cell fusion was observed for the Sf9 cells transfected with the baculovirus expressing NSvc2-N or with the empty baculovirus ( Fig 6C and 6D ) . In addition , cell-cell membrane fusion was observed for the Sf9 cells co-infected with the baculovirus expressing NSvc2-N and the virus expressing NSvc2-C ( Fig 6F and 6G ) , confirming that only RSV NSvc2-C played a role in cell-cell membrane fusion under the acidic condition . To further investigate the function of NSvc2-C in cell-cell membrane fusion , we generated a three-dimensional ( 3D ) model structure of NSvc2-C through a homology-based modeling approach ( S6A and S6B Fig ) . The 3D model of NSvc2-C consisted of three distinct domains: domain I ( yellow ) , II ( red ) , and III ( blue ) . Two putative fusion loops ( green ) were predicted at the top of the NSvc2-C model ( S6B Fig ) . Similar loops were also reported to be responsible for cell-cell membrane fusion during animal virus infections [49 , 50] . We then constructed three NSvc2-C mutants with one or two fusion loops ( FL ) deleted ( i . e . , ΔFL1 , ΔFL2 and ΔFL1+ΔFL2 ) . These mutants were infected individually into Sf9 cells and their expressions were confirmed individually by immunoblotting assays ( S6D and S6E Fig ) . Fusogenic activities of the three deletion mutants were then examined in the Sf9 cells using the recombinant baculovirus expression system . Results showed that the number of syncytial cells induced by the ΔFL1 or ΔFL2 mutant was much less than that induced by the WT NSvc2-C . When the ΔFL1 + ΔFL2 mutant was expressed in the Sf9 cells , the number of syncytial cells was further decreased ( S6F and S6G Fig ) . Based on this finding , we concluded that both fusion loops of NSvc2-C played important roles in cell-cell membrane fusion . To identify the amino acid residue ( s ) important for the fusogenic activity , we aligned the RSV fusion loop sequences ( Cys459-Cys465 [Loop 1] and Cys485-Tyr498 [Loop 2] ) with the loop sequences from other four tenuiviruses ( Fig 6H ) . The alignment result indicated that the two fusion loop sequences were relatively conserved among the five tenuiviruses . Six conserved hydrophobic amino acid residues ( Phe460 , Phe489 , Phe492 , Tyr494 , Pro496 and Tyr498 ) were found in a unique vertex in the modeled NSvc2-C 3D structure ( Fig 6I ) . Introduction of mutations into each of the six amino acid residues showed that three of them ( F460A , F489A , and Y498A ) were important for the membrane fusion activity ( Fig 6J and 6K ) . The above results showed that RSV NSvc2-N and NSvc2-C had different functions during RSV acquisition by SBPH . To elucidate the role of NSvc2-C in RSV entrance into midgut cells during virus acquisition , we developed a new system to analyze the wild type NSvc2 ( NSvc2-WT ) and its fusion loop mutants for their abilities to release RSV virions from endosomes . In this system , purified RSV virions were incubated with a cell extract from Sf9 expressing NSvc2-WT for 3 h and then used it to feed SBPHs for 24 h . Midguts were isolated from the SBPHs and probed for the presence of RSV virions and NSvc2-N through immuno-labeling . Result showed that RSV virions ( green ) and NSvc2-N ( red ) were both present in the endosomal-like vesicles inside epithelial cells by 24 h post feeding ( Fig 7A , upper row ) , indicating that both RSV virions and NSvc2-N had entered the midgut cells using this system . We then incubated purified RSV virions with cell extracts from Sf9 expressing the NSvc2N114A/N199A/N232A or NSvc2F460A/F489A/Y498A mutant and used them individually to feed SBPHs . The result showed that RSV virions were detected in the midgut lumen , but not in the epithelial cells , of the SBPHs fed with the mixture containing RSV virions and cell extract with the NSvc2N114A/N199A/N232A mutant ( Fig 7A , middle row ) . The result also showed that RSV virions were present in the epithelial cells and accumulated inside the endosomal-like structures of the SBPHs fed with the mixture containing RSV virions and cell extract with the NSvc2F460A/F489A/Y498A mutant . In addition , more endosomal-like structures with RSV virions were found in the epithelial cells of the SBPHs fed with the mixture containing RSV virions and the NSvc2F460A/F489A/Y498A mutant than that in the epithelial cells of the SBPHs fed with the mixture containing RSV virions and NSvc2-WT ( Fig 7A , bottom row ) . To investigate the effects of recombinant NSvc2-WT or its mutants on the acquisition and transmission of RSV by SBPH using the new developed system , the NSvc2-WT or its mutants were expressed from Sf9 cells , mixed with purified RSV virions , and then used to feed SBPHs . The results indicated that the SBPHs fed with the mixture containing RSV virions and NSvc2-WT had 12–18% RSV acquisition rate and 4–5% virus transmission rate ( Fig 7B and S3 Table ) . For the SBPHs fed with the mixture containing RSV virions and the NSvc2F460A/F489A/Y498A mutant , only 2–3% RSV acquisition rate and 0–1% virus transmission rate were observed . No RSV acquisition and transmission were found for the SBPHs fed with the mixture containing RSV virions and the NSvc2N114A/N199A/N232A mutant . In the subsequent experiments , SBPHs were first fed with NSvc2-WT or its mutants and then RSV virions followed by the RSV acquisition and transmission assays . The results showed that the SBPHs fed with NSvc2-WT first and then RSV virions had 15–20% RSV acquisition rate and 6–8% virus transmission rate ( Fig 7C and S4 Table ) . Only 4–5% RSV acquisition rate and 1–2% virus transmission rate were observed for the SBPHs fed with the NSvc2F460A/F489A/Y498A mutant first and then RSV virions ( Fig 7C and S4 Table ) . This result indicated that feeding SBPHs with NSvc2-WT first and then RSV virions could cause slightly higher virus acquisition and transmission rate compared with the SBPHs fed with the mixture containing both RSV virions and NSvc2-WT . Again , no RSV acquisition and transmission were observed for the SBPHs fed with the NSvc2N114A/N199A/N232A mutant first and then RSV virions ( Fig 7C and S4 Table ) . In a separate experiment , purified RSV virions were micro-injected into the hemocoel of SBPHs followed by RSV acquisition and transmission assays . The result showed that the micro-infected SBPHs had 66–72% RSV acquisition rate and 22–27% virus transmission rate ( S5 Table ) , indicating that the midgut barrier ( s ) could be bypassed by micro-injecting RSV virions directly into SBPH hemocoel . Thus , the fusogenic activity deficient NSvc2 mutant is capable of passaging RSV virions through epithelial cells but is defective in releasing RSV virions from the endosomal-like vesicles .
Understanding how plant viruses are transmitted by their insect vectors is one of the key steps to manage virus diseases worldwide . Insect midgut is a major barrier to block the entrance of non-compatible plant viruses into insect vector . In this study , we used RSV and its SBPH vector as a working model to elucidate the molecular mechanism controlling RSV virion entrance into SBPH midgut cells for transmission . Through various assays , we have now determined that RSV , a circulative and propagative transmitted tenuivirus , utilizes its glycoprotein NSvc2 as a helper component to ensure the successful entrance of virions into the midgut of SBPH . This is the first evidence showing that tenuivirus uses a helper component mediated mechanism for persistent-propagative virus transmission . Although the function of RSV glycoprotein NSvc2 during virus acquisition and transmission was previously proposed to be similar to the glycoproteins of other plant-infecting tospoviruses or animal-infecting bunyaviruses [31 , 51 , 52] , our findings from this study showed that , unlike tospoviruses and bunyaviruses , RSV NSvc2 is not present on the surface of RSV non-enveloped filamentous virions ( Fig 2B and Fig 3D–3F ) , and purified RSV virions were unable to enter SBPH midgut cells . The inability of RSV virions to enter the midgut cells could , however , be rescued by the addition of NSvc2 ( Fig 2D and Fig 7 ) . In addition , micro-injection of RSV virions directly into SBPH hemolymph could bypass the midgut barriers ( S5 Table ) [43] . More importantly , in the presence of NSvc2 , purified RSV virions could be successfully acquired by SBPHs and then transmitted to rice seedlings ( Fig 2C , Fig 7B and 7C ) . For non-persistent , semi-persistent , and persistent-nonpropagative transmitted plant viruses , the helper component proteins or the helper factors were shown to mediate the interactions between virus virions and their insect vectors [4 , 6 , 15 , 17] . Although these helper component proteins or factors are not located on the surface of purified virions , they are absolutely required for insect transmissions [20 , 53] . The results presented in this paper indicate that this previously proposed helper component theory can also be applied to explain the function of RSV NSvc2 during virus circulative and propagative transmission . First , the RSV NSvc2 is not a virion structural protein ( Fig 2B and Fig 3 ) . Second , NSvc2 can interact with RSV virions and bind to SBPH midgut ( Fig 2E , and 2F and Fig 3 ) . Third , NSvc2 is absolutely required for the entrance of RSV virions into SBPH midgut cells ( Fig 2 and Fig 7 ) . Moreover , the SBPHs fed first with the recombinant NSvc2 and then purified RSV virions had slightly higher RSV acquisition rate and virus transmission rate compared with the SBPHs fed with the recombinant NSvc2 and purified RSV virions simultaneously ( Fig 7C ) . Consequently , we consider RSV NSvc2 as a helper component that is needed for the successful overcome of midgut barrier ( s ) in SBPH for virus persistent-propagative transmission . This helper component may function as a molecular bridge to ensure the proper interaction between RSV virions and SBPH midgut receptor ( s ) . We , however , cannot rule out other alternative possibilities . Neither can we rule out the hypothesis that tenuiviruses can form membrane bound complexes with NSvc2 , and these complexes are unstable during virus purification . However , earlier electron microscopic results using fixed tissue sections from RSV-infected rice leaves or RSV-infected SBPH tissues did not reveal the presence of membrane bound complexes in RSV-infected cells [54 , 55] . RSV NSvc2 is known to be processed into NSvc2-N and NSvc2-C in the RSV-infected cells [41 , 42] . We have found that NSvc2-N accumulated on the midgut surface during SBPH acquisition of RSV . The ectopically expressed soluble NSvc2-N:S could bind directly to the midgut surface and inhibit the subsequent RSV acquisition by SBPH . Our enzymatic deglycosylation results showed that NSvc2-N could be modified by N-linked but not O-linked glycans . The glycan-modification of NSvc2-N might be different from the N- or O-glycosylation of TSWV Gn protein reported previously [30] . It is noteworthy that the N-glycosylation deficient NSvc2-NN114A/N199A/N232A mutant was unable to interact with midgut surface receptor ( s ) and was unable to block RSV acquisition and transmission by SBPH . In contrast to the recombinant NSvc2-WT purified from the Sf9 cells , the N-glycosylation deficient NSvc2 mutant was unable to mediate the entrance of RSV virions into SBPH midgut cells . The mutant failed to cause RSV acquisition and virus transmission by SBPH ( Fig 7 ) . Based on these results , we propose that N-glycosylation modification has an important role in the interaction between RSV NSvc2 and SBPH midgut surface . Sugar transporter 6 protein was recently reported to play a critical role in RSV invasion of SBPH midgut epithelial cells [56] . This sugar transporter interacted with RSV NP and was highly expressed in the midgut cells of SBPH . Expression of this protein in Sf9 cells allowed RSV virions to enter the cells [56] . In our study , RSV virions were detected only in the midgut lumen and not inside the epithelial cells of the SBPHs fed with purified RSV virions ( Fig 2D , row 3 ) . In the presence of NSvc2 , however , RSV virions did enter the epithelial cells ( Fig 2D and Fig 7 ) , supporting the conclusion that NSvc2 is absolutely required for the passage of RSV virions through SBPH midgut barrier ( s ) . We speculate that SBPH midgut may have more barriers for RSV virions entrance than that of the Sf9 cells . The outer capsids of rice black streak dwarf virus ( RBSDV ) and southern rice black streak dwarf virus ( SRBSDV ) were also shown to interact with SBPH sugar transporter 6 protein [56] . It would be interesting to see if NSvc2 can interact with SBPH sugar transporter 6 protein and if glycosylation of NScv2 plays an important role in the interaction between the two proteins . Genomic RNAs of many animal-infecting viruses are encapsidated within bilayer lipid envelopes , which protect viruses during their transmissions between hosts . For bunyaviruses , the glycoprotein Gn and Gc on surface membrane envelopes are known to interact with viral ribonucleoprotein complexes ( RNPs ) which composed of the viral genomic RNAs and the nucleocapsid protein inside virions [51 , 57] . Although the purified RSV virions are neither enveloped nor carrying the glycoprotein NSvc2 ( Fig 3 ) , the NSvc2 protein is associated with RSV virions during the entrance of virus into SBPH midgut . Using Yeast-Two-Hybrid and co-IP analysis , RSV glycoprotein NSvc2-N and NSvc2-C were also found to interact with RSV virions . These interactions are likely to form a stable RSV virion:NSvc2-N:NSvc2-C complexes during virus transmission by its SBPH vector . Endocytosis was shown to be important for the entrance of animal-infecting bunyaviruses into host cell [51 , 58 , 59] . Once inside a new host , the viral lipid envelopes fused with host cell membranes to release viral RNPs into the cytoplasm to start an infection [49 , 58 , 60] . After the recognition of RSV virions , midgut epithelial cells underwent endocytosis to compartmentalize RSV virions:NSvc2-N:NSvc2-C complexes in the early and then the late endosomes ( Fig 5 ) . After that , RSV virions were released from the late endosomes into the cytosol for further virus replication and spread into adjacent cells . Only RSV virions:NSvc2-N complexes were released into the cytosol , while the NSvc2-C was remained inside the endosomes . Because tenuivirus virions are not enveloped , it is unclear how membrane fusion affects the release of RSV virions into the cytosol . It was reported that acidic condition inside the late endosomes could trigger the conformation of bunyavirus Gc protein and induce membrane fusion that could disrupt the enveloped structure of virus virions and lead to the release of viral RNPs into the cytoplasm of host cell [50 , 61] . It is possible that under the acidic condition , NSvc2-C also changes its structure and stops its association with RSV virions . Meanwhile , the membrane fusion triggered by NSvc2-C releases the RSV virions into the host cell cytosol . Bunyavirus glycoprotein Gc was reported to be activated and to cause membrane fusion under the acidic conditions inside endosomes [62 , 63] . We have also determined that under the acidic condition , NSvc2-C could cause Sf9 cell membrane to fuse ( Fig 6 ) . Crystal structures of several animal-infecting bunyavirus glycoprotein Gc have been determined and are considered as class II fusion proteins [61 , 64 , 65] . For the class II fusion proteins , the hydrophobic fusion loops were reported to be at the apex of domain II [64 , 66] . For all known viral fusion proteins , the fusion loops are the key motifs that are inserted into the target membranes to bring the virus and target membranes together . Using a homology-based modeling approach , RSV NSvc2-C was also found to have a class-II fusion glycoprotein architecture with two putative fusion loops . Deletion of the two fusion loops resulted in a defect in membrane fusion . The two fusion loops found in NSvc2 are highly conserved among the members in the genus Tenuivirus , suggesting that they have conserved roles in membrane fusion . With the aid of homology-based modeling , we have determined that residue F460 , F489 and Y498 in the fusion loop of NSvc2-C played critical roles in cell membrane fusion . Also , the NSvc2 mutant that failed to cause cell membrane fusion was unable to release RSV virions from endosomes into the cytosol . This mutant also caused much lower RSV acquisition rate and virus transmission rate by SBPH . Current gene function studies on plant multi-segmented negative-strand RNA viruses are difficult , due mainly to the lack of proper reverse genetics methods . In this study , we developed a new method to overcome this obstacle . We first expressed the WT or mutant NSvc2 in Sf9 insect cells , and then incubated the extracts from the infected Sf9 cells with purified RSV virions . After feeding SBPHs with a mixture containing purified RSV virions and NSvc2 or a mixture containing purified RSV virions and a mutant NSvc2 , we have determined that in the presence of NSvc2 , purified RSV virions were able to overcome the midgut barriers to enter earlier and late endosomes in the epithelial cells , and then be released from late endosomes to the cytosol for further virus replication and transmission . We have also been able to analyze the function of NSvc2 mutants during virus entrance to SBPH midgut cells . This new method should benefit gene function studies for viruses whose infectious clones are currently difficult to make . We think that this technology can not only be used to investigate the functions of tenuivirus glycoproteins but also the functions of glycoproteins encoded by other plant multi-segmented negative-strand RNA viruses . Taken together , we conclude that the circulative and propagative transmitted RSV uses a helper component strategy to ensure the entrance of RSV virions into SBPH midgut during vector transmission . Based on the findings presented in this paper , we propose a working model for deciphering how plant-infecting tenuiviruses overcome the SBPH midgut barriers ( Fig 8 ) . In this model , plant sap containing RSV virions , NSvc2-N , and NSvc2-C is acquired into the midgut lumen during SBPH feeding on the RSV-infected rice plants . The NSvc2-N protein recognizes the unidentified midgut cell surface receptor ( s ) and acts as a helper component to ensure the interaction between RSV virions and midgut surface receptor ( s ) . Upon attachment of RSV virions:NSvc2-N:NSvc2-C complexes to the midgut surface receptor ( s ) , midgut cells undergo endocytosis , resulting in compartmentalization of RSV virions:NSvc2-N:NSvc2-C complexes in the early and then in the late endosomes . The acidic condition inside the late endosomes triggers a conformation change of NSvc2-C , and the conformation-changed NSvc2-C causes the membrane fusion . Finally , the RSV virions:NSvc2-N complexes are released from endosomes into the cytosol . The findings presented in this paper revealed a new type of virus–insect midgut interaction that requires a virally encoded helper component during virus persistent-propagative transmission .
Rice stripe virus was previously isolated from an RSV-infected rice plant and then maintained inside a growth chamber at the Jiangsu Academy of Agricultural Sciences , Nanjing , Jiangsu Province , China . Small brown planthopper ( SBPH ) was reared on rice seedlings cv . Wuyujing NO . 3 inside growth incubators set at 26 . 5°C and a photoperiod of 16 h / 8 h ( light / dark ) . Rice seedlings were changed once every 12 days as described [67] . Eight hundred milliliter precooled 0 . 1 M phosphate buffer ( PB ) , pH 7 . 5 , with 0 . 01 M EDTA was added to 50-gram RSV-infected rice leaf tissues followed by 5 min homogenization in a blender . The homogenate was centrifuged at 8 , 000 × g for 20 min at 4°C . The resulting supernatant was mixed with PEG 6000 ( 6% ) and NaCL ( 0 . 1 M ) , and then stirred overnight at 4°C . After centrifugation at 8 , 000 × g for 20 min , the pellet was resuspended in 0 . 01 M PB followed by 2 h centrifugation at 150 , 000 × g . The pellet was resuspended in 6 ml 0 . 01 M PB and laid on the top of a 4 ml 20% glycerol cushion inside a centrifuge tube followed by a centrifugation at 150 , 000 × g for 2 h . Different fractions inside the centrifuge tube were carefully collected and the pellet was resuspended in 4 mL PB buffer containing 30% glycerol . The collected samples were stored at -70°C till use . Immunofluorescence labeling was performed as described previously with specific modifications [45] . Midguts were obtained from second-instar nymphs and fixed overnight in a 4% paraformaldehyde ( PFA ) ( Thermo Fisher Scientific , Waltham , MA USA ) solution at 4°C . After three rinses in a 0 . 01 M phosphate-buffered saline ( PBS ) , pH 7 . 4 , the midguts were treated for 30 min in a 2% Triton X-100 solution followed by 1 h incubation in a 1:200 ( v/v ) diluted specific primary antibody . The midguts were then incubated in a 1:200 ( v/v ) diluted specific fluorescence conjugated secondary antibody for 2 h at room temperature ( RT ) . The midguts were rinsed three times in the PBS and then mounted in an antifade solution ( Solarbio , Shanghai , China ) . The mounted midguts were examined under an inverted Leica TCS SP8 fluorescent confocal microscope ( Leica Microsystems , Solms , Germany ) . Images of single sections were taken to show the co-localizations of two assayed proteins . The overlapped coefficient value was determined using the LAS X software , and 50 merged spots from 30 guts per experiment were analyzed . Rabbit polyclonal antibodies against RSV NSvc2-N or NSvc2-C were produced in our laboratory . Mouse monoclonal antibody against RSV NP was a gift from Professor Jianxiang Wu , Zhejiang University , Hangzhou , China . Rab5 anti-rabbit IgG and Rab7 anti-rabbit IgG were from the Cell Signaling Technology ( Danvers , MA , USA ) , and EEA1 anti-mouse IgG was from Novus ( Centennial , CO , USA ) . Secondary antibodies used in this study were FITC conjugated rabbit anti-mouse IgG ( F9137 ) or goat anti-rabbit IgG ( F9887 ) , and TRITC conjugated goat anti-rabbit IgG ( T6778 ) from Sigma-Aldrich ( St . Louis , MO , USA ) . Alexa Fluor 647 phalloidin ( A22287 ) and Rhodamine Phalloidin ( R415 ) were from Invitrogen ( Carlsbad , CA , USA ) . These antibodies were used at 1:200 ( v/v ) dilution . Formvar/carbon-coated nickel grids ( 200 mesh ) were floated individually on drops of PB buffer containing purified RSV virions for 2 min . The grids were then transferred onto drops of 1% BSA buffer and incubated for 10 min followed by 1 h incubation on drops of 1:200 ( v/v ) diluted mouse monoclonal antibody against RSV NP or rabbit polyclonal antibody against NSvc2-N . The grids were rinsed with several drops of PB and then incubated on drops of 1:30 ( v/v ) diluted goat anti-mouse IgG conjugated with 15 nm gold particles or goat anti-rabbit IgG conjugated with 8 nm gold particles for 1 h . After several rinses with PB , the grids were stained with 2% uranyl acetate and then examined under a Hitachi HT-7700 transmission electron microscope . Preparation of recombinant baculoviruses was as described previously [68] . Sequence encoding RSV NSvc2 , NSvc2-N , NSvc2-C , NSvc2-N:S ( amino acid position 20 to 265 , lacking the signal peptide and the transmembrane domain ) , and TSWV Gn:S was PCR-amplified using a cDNA made from an RSV-infected rice plant or from a TSWV-infected Nicotiana benthamiana plant using specific primers ( S6 Table ) . The NSvc2-N and NSvc2-C contained a FLAG tag and the NSvc2-N:S contained a six-His tag sequence at their 3' end . A Gp64 signal peptide sequence was then fused to the 5' end of RSV NSvc2 or TSWV Gn:S proteins via overlapping PCR and the product was cloned into vector pFastBac1 ( S6 Table ) . The site-directed mutagenized mutants were constructed as described [69] . All plasmids were sequenced and then transformed individually into DH10BacTM cells to generate recombinant baculoviruses as instructed by the manufacturer ( Invitrogen ) . The recombinant baculoviruses were mixed individually with the FnGENE HD Transfection Reagent ( Promega ) and infected Sf9 cells as instructed . Sf9 cells were infected with various recombinant baculoviruses at the multiplicity of infection ( MOI ) of five . The infected Sf9 cells were collected after 72 h incubation and were lysed in 10 ml PBS buffer using an ultrasonic cell crusher followed by centrifugation at 4°C to remove cell debris . Supernatant from individual treatment was collected , incubated with a nickel-nitrilotriacetic acid resin ( Ni-NTA , Germany ) for 2 h , and then loaded onto a chromatographic column ( Bio-Rad Hercules , California , USA ) . After separation , the column was washed with two bed volumes of 50 mM imidazole in PBS , and the recombinant proteins were then eluted from the columns with 250 mM imidazole solution followed by dialysis against PBS . The purified recombinant proteins were stored at -70°C until use . Purified NSvc2-N:S and other mutant proteins were individually deglycosylated with PNGaseF to remove the N-linked glycans or with Neuraminidase and O-Glycosidase to remove the O-linked glycans as instructed ( New England Biolabs , Ipswich , MA , USA ) . Sodium dodecyl sulfate ( SDS ) and dithiothreitol ( DTT ) were added to the purified protein samples , respectively , and the mixed samples were heated at 100°C for 10 min . After denaturation , buffer and glycosidases were added to the samples followed by 3 h incubation at 37°C . The enzyme-treated samples were mixed with a loading buffer containing SDS and boiled for 10 min prior to electrophoresis in 10% ( w/v ) SDS-PAGE gels . The separated proteins were transferred onto nitrocellulose membranes and the membranes were probed individually with the NSvc2-N antibody diluted at 1:5 , 000 ( v/v ) followed by a goat anti-rabbit IgG HRP conjugate ( 31466 , Invitrogen ) diluted at 1:10 , 000 ( v/v ) . The detection signal was visualized using the ChemiDoc Touch Imaging System as instructed ( Bio-Rad ) . Midgut binding assays were performed as described previously [30] . Briefly , second-instar nymphs were placed inside open-ended Eppendorf centrifuge tubes and fed with purified glycoproteins resuspended in a TF buffer ( PBS with 10% glycerol , 0 . 01% Chicago sky blue , and 5 mg / ml BSA ) through a stretched parafilm membrane . After 3 h feeding , the glycoprotein samples were replaced with a 10% sucrose solution and the insects were allowed to feed for another 12 h to clear their midguts , indicated by the disappearance of the Chicago sky blue dye from the midguts . The insects were then dissected , fixed in a 4% paraformaldehyde solution and analyzed for the binding through immunofluorescence labeling as described above . Second-instar nymphs were placed inside empty bottles for 2 h and then allowed to feed on purified recombinant NSvc2-N:S protein or mutant proteins through stretched parafilm membranes for 24 h . The pre-fed SBPHs were then allowed to feed on RSV-infected rice seedlings for 48 h . Second-instar nymphs were also fed on a purified RSV virion , supernatanta mixture of both or mixture of purified RSV virions and Sf9 cell extract containing NSvc2 proteins through stretched parafilm membranes for 24 h followed by 12 day feeding on healthy rice seedlings . After that , the SBPHs were individually transferred onto individual healthy rice seedlings cv . Wuyujing NO . 3 inside glass tubes for virus transmission test for 24 h . After virus transmission , the seedlings were allowed to grow for 14 days prior to RSV infection test . One hundred SBPHs and one hundred rice seedlings were used for each treatment . In a separate experiment , purified RSV virions ( about 70 ng/μl ) were micro-injected into the hemocoel of SBPHs using a microprocessor-controlled Nanoliter 2010 injector ( World Precision Instruments , Sarasota , FL , USA ) . The RSV acquisition and virus transmission of the injected SBPHs were performed using the same methods as above . The rates of RSV acquisition and transmission by SBPHs were determined by immuno-dot blot assays . SBPHs were collected individually at 12 days post 48 h feeding on the RSV-infected rice seedlings or on a purified RSV virion sample , or at 14 days post 24 h feeding on the purified RSV viroins , recombinant protein or a mixture of both . The collected SBPHs were grinded individually in PBS buffer . The samples were centrifuged at 5 , 000 rpm for 1 min at 4°C to remove cell debris and the supernatant from each sample was blotted onto nitrocellulose membranes . The membranes were probed with a 1:5 , 000 ( v/v ) diluted RSV NP antibody and then with a 1:20 , 000 diluted secondary alkaline phosphatase ( AP ) -conjugated goat anti-mouse IgG ( Sigma ) . The detection signal was visualized using a 5-bromo-4-chloro-3-indo-lylphosphate-nitroblue tetrazolium ( BCIP-NBT ) solution ( Sangon Biotech , Shanghai , China ) . Three independent experiments with 50 nymphs per treatment were performed . Total RNA was isolated from individual SBPHs using TRIzol ( Thermo Fisher Scientific ) . cDNAs were synthesized using the M-MLV reverse transcriptase ( M1701 , Promega ) . Quantitative PCR was performed on a CFX Connect Real-Time System ( Bio-Rad , USA ) using the PowerUp SYBR Green Master Mix ( A25742 , Thermo Fisher Scientific ) . RSV specific primers were designed using the Primer Premier 5 . 0 software . The expression level of SBPH Actin gene was used as an internal reference . The relative abundance of RSV RNA in SBPHs was calculated using the 2 –ΔΔCt method . Primers used for the RT-qPCR are listed in S6 Table . The coding sequence of RSV NSvc2-N or NSvc2-C was cloned into a bait vector pGBKT7 . The coding sequence of RSV NP was cloned into a prey vector pGADT7 . Primers used in PCR reactions are listed in S6 Table . Yeast two-hybrid assays were performed as previously described [70 , 71] . Briefly , a bait vector and a prey vector ( see details in the result section ) were co-transformed into the Y2HGold strain cells by heat shock method . In addition , vector pGADT7-T was co-transformed with vector pGBKT7-53 or vector pGBKT7-Lam into the Y2HGold strain cells and the cells were used as the positive and the negative control , respectively . All the cells were first grown on a synthetic dextrose medium lacking Tryptophan and Leucine amino acid ( SD-Trp-Leu ) for 3 days at 30°C , and then on a synthetic dextrose medium lacking Tryptophan , Leucine , Histidine and Ademethionine amino acid ( SD-Trp-Leu-His-Ade ) for 5 days at 30°C . Sf9 cells were transfected with different recombinant baculoviruses at a MOI of five . After 48 h infection , the infect cells were washed twice with a fresh medium , treated in a PBS buffer , pH 5 . 0 , for 2 min , and then grown in a pH neutral medium for 4 h at 28°C . The cell was then examined for cell-cell membrane fusion under light/bright bield ofan Olympus IX71 inverted microscope ( Olympus , Hamburg , Germany ) . Homology modeling of NSvc2-C was as described previously [72 , 73] . Briefly , RSV NSvc2-C sequence was used to search the I-TASSER Server ( https://zhanglab . ccmb . med . umich . edu/I-TASSER/ ) . Based on the high TM-score value ( 0 . 805 ) , glycoprotein of Sever Fever with Thrombocytopenia Syndrome Phlebovirus ( SFTSV ) ( PDB: 5G47 ) was chosen as the template to build the homology-based model of RSV NSvc2-C . Amino acid residues and their surface positions in the three-dimensional structure were predicted using the PyMOL program ( https://pymol . org/2/ ) .
|
Over 75% of the known plant viruses are insect transmitted . Understanding how plant viruses interact with their insect vectors during virus transmission is a key step towards the successful management of plant viruses worldwide . Several models for the direct or indirect virus–insect vector interactions have been proposed for the non-persistent or semi-persistent virus transmissions . However , the mechanisms controlling the interactions between viruses and their insect vector midgut barriers are poorly understood . In this study , we demonstrated that the circulative and propagative transmitted rice stripe virus ( RSV ) utilized its glycoprotein NSvc2 as a helper component to ensure a specific interaction between its virions and SBPH midgut cells to overcome the midgut barriers inside this vector . This is the first report of a viral helper component mediated mechanism for persistent-propagative virus transmission . Our new findings and working model should expand our knowledge on the molecular mechanism ( s ) controlling the interaction between virus and its insect vector during virus circulative and propagative transmission in nature .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"vesicles",
"microbiology",
"viral",
"structure",
"epithelial",
"cells",
"membrane",
"fusion",
"rice",
"experimental",
"organism",
"systems",
"seedlings",
"plants",
"cellular",
"structures",
"and",
"organelles",
"insect",
"vectors",
"endosomes",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"grasses",
"animal",
"cells",
"animal",
"studies",
"biological",
"tissue",
"cell",
"membranes",
"virions",
"disease",
"vectors",
"eukaryota",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"anatomy",
"virology",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"species",
"interactions",
"cytosol",
"organisms"
] |
2019
|
Tenuivirus utilizes its glycoprotein as a helper component to overcome insect midgut barriers for its circulative and propagative transmission
|
Human hookworms ( Necator americanus , Ancylostoma duodenale , and Ancylostoma ceylanicum ) are intestinal blood-feeding parasites that infect ~500 million people worldwide and are among the leading causes of iron-deficiency anemia in the developing world . Drugs are useful against hookworm infections , but hookworms rapidly reinfect people , and the parasites can develop drug resistance . Therefore , having a hookworm vaccine would be of tremendous benefit . We investigated the vaccine efficacy in outbred Syrian hamsters of three A . ceylanicum hookworm antigen candidates from two classes of proteins previously identified as promising vaccine candidates . These include two intestinally-enriched , putatively secreted cathepsin B cysteine proteases ( AceyCP1 , AceyCPL ) and one small Kunitz-type protease inhibitor ( AceySKPI3 ) . Recombinant proteins were produced in Pichia pastoris , and adsorbed to Alhydrogel . Recombinant AceyCPL ( rAceyCPL ) /Alhydrogel and rAceySKPI3/Alhydrogel induced high serum immunoglobulin G ( IgG ) titers in 8/8 vaccinates , but were not protective . rAceyCP1/Alhydrogel induced intermediate serum IgG titers in ~60% of vaccinates in two different trials . rAceyCP1 serum IgG responders had highly significantly decreased hookworm burdens , fecal egg counts and clinical pathology compared to Alhydrogel controls and nonresponders . Protection was highly correlated with rAceyCP1 serum IgG titer . Antisera from rAceyCP1 serum IgG responders , but not nonresponders or rAceyCPL/Alhydrogel vaccinates , significantly reduced adult A . ceylanicum motility in vitro . Furthermore , rAceyCP1 serum IgG responders had canonical Th2-specific recall responses ( IL4 , IL5 , IL13 ) in splenocytes stimulated ex vivo . These findings indicate that rAceyCP1 is a promising vaccine candidate and validates a genomic/transcriptomic approach to human hookworm vaccine discovery .
Human hookworms ( Necator americanus , Ancylostoma duodenale , and Ancylostoma ceylanicum ) are soil-transmitted nematodes ( STNs ) that infect the small intestine and feed on blood [1] . Human STNs encompass three phylogenetically distant parasites: hookworms , large roundworms ( Ascaris lumbricoides ) , and whipworms ( Trichuris trichiura ) . Among human STNs , hookworms carry the highest disease burden [2] . In children , infection by hookworms causes significant growth stunting , cognitive deficiencies , malnutrition , iron-deficiency anemia and hypoproteinemia; in adults infection results in adverse birth outcomes ( e . g . , low birthweight babies ) and reduced productivity [3–5] . It is estimated that hookworms infect ~500 million people worldwide [6] , and although concentrated in Latin America , sub-Saharan Africa , and Southeast Asia , even people in impoverished regions of the United States ( US ) still get infected [1 , 7] . Once at 40% prevalence in the southern US ( circa 1911 ) , hookworm infections led to an estimated 43% reduction in future earnings of children infected and were responsible for 22% of the income gap and 50% of the literacy gap between North and South [8] . The elimination of hookworms via treatment campaigns , improved sanitation , education , and economic development undoubtedly had a major impact on the vitalization and success of the South today . Currently , hookworm disease is estimated to cause 4 . 1 million disability adjusted life years ( DALYs ) and US$139 billion in indirect economic losses each year [9] . Hookworms are the second most important parasitic cause of global anemia after malaria [10] . Mass drug administration ( MDA ) of benzimidazoles ( albendazole , mebendazole ) in school-aged children is the current control measure for hookworms [11 , 12] . From 1990–2013 , MDA reduced hookworm prevalence by only 5 . 1% , compared to 25 . 5% reduction for A . lumbricoides [6] . Additionally , poor efficacy of mebendazole against hookworms is well known [13] and poor or reduced efficacy of albendazole against hookworms is being reported in multiple locations around the world ( e . g . , egg reduction rates as low as 0% in Ghana [14] and cure rates as low as 36% in Lao PDR [15] ) . Veterinary parasites phylogenetically and biologically similar to hookworms ( e . g . , the blood-feeding Haemonchus contortus ) develop resistance to anthelmintic drugs frequently , rapidly , and broadly [16 , 17] . Water , sanitation and hygiene ( WASH ) is being explored as a control strategy to combine with MDA [18] . Improvements in WASH ( water , sanitation , and hygiene ) , although important , are insufficient to tackle the enormous STN problem alone [18–20] . Having a vaccine to prevent infection from occurring in the first place would be of tremendous benefit . Although it is widely accepted that a hookworm vaccine is needed [21] , there is only a single phase 1 clinical trial underway testing two individual recombinant hookworm proteins formulated on the Th2 adjuvant Alhydrogel [22] . There are no other candidates in advanced preclinical development [23] . Because targeting infectious third staged larval ( L3i ) antigens carries the risk of triggering allergic reactions in previously exposed people [24] , efforts are focused on adult stage antigens , namely an aspartic protease ( APR1 ) and a glutathione S-transferase ( GST1 ) [25] . Both APR1 and GST1 localize to the adult canine hookworm Ancylostoma caninum intestine ( and non-intestinal tissues ) ; these enzymes are thought to help digest hemoglobin and detoxify heme , respectively [25–28] . In canine and hamster models , recombinant protein immunogens from A . caninum gave 33–53% decreased hookworm burdens [26 , 27 , 29] . However , in a phase 1a trial , although rNaGST1 was safe and immunogenic in human volunteers [30] , IgG had negligible neutralizing effect on rNaGST1 enzymatic activity , despite the observation that IgG from immunized mice were highly neutralizing [31]; these results suggest a decreased potential for this vaccine in human trials . It remains crucial that further and expanded efforts be undertaken to develop new hookworm vaccines . Development of hookworm vaccines , however , have been limited and lag far behind more concerted efforts , such as against malaria [23 , 32] . This may be in part because full genomes for human hookworms were formerly unavailable , which prevented large-scale reverse vaccinology [33] against hookworms . This situation has recently been addressed for all three species of human hookworms [34–36] . Genomics and transcriptomics for A . ceylanicum hookworm infections in hamsters is being used to identify new and potent vaccine antigen candidates [35] . Syrian hamsters are the only laboratory rodent permissive for the human hookworm life cycle , and A . ceylanicum infections in hamsters are an excellent model for hookworm infections in humans [37] . We previously identified two classes of A . ceylanicum genes , that are strongly expressed and upregulated during blood feeding , as encoding potential antigen candidates [35]: cathepsin B cysteine proteases ( CPs ) and small Kunitz-type protease inhibitors ( SKPIs ) . Here , we explore the vaccine efficacy of two CPs and one SKPI ( AceyCP1 , AceyCPL , AceySKPI3 ) using the A . ceylanicum hookworm—hamster model system , and investigate functional and immunological aspects of protection with one of these vaccine candidates .
Animal experimentation was carried out under protocols approved by the University of Massachusetts Medical School Institutional Animal Care and Use Committees ( IACUC ) . All housing and care of laboratory animals used in this study conform to the NIH Guide for the Care and Use of Laboratory Animals in Research and all requirements and all regulations issued by the United States Department of Agriculture , including regulations implementing the Animal Welfare Act ( P . L . 89–544 ) . Male Syrian hamsters ( strain HsdHan:AURA ) were purchased from Envigo at 3 weeks of age and housed 4 hamsters per cage . Hamsters were provided with food and water ad libitum . Male hamsters were used for all studies because female hamsters are ~5-fold less susceptible to A . ceylanicum infection . Using females would require much larger numbers of animals to achieve adequate infection intensity , prevalence , and statistical significance . To assess intestinal versus non-intestinal gene expression for A . ceylanicum genes ( including those encoding vaccine candidates in this study ) , we used RNA-seq data for A . ceylanicum that we and others had previously generated from whole worms and adult male intestine [35 , 38] . These data were a mixture of paired- and single-end reads with varying lengths . To make cross-comparisons of these data as unbiased as possible , we trimmed all read sets to have single-end 50-nt reads , using quality_trim_fastq . pl and the arguments "-q 33 -u 50" . We quality-filtered RNA-seq reads by running Trimmomatic 0 . 36 with the following arguments: "java -jar $TRIM/trimmomatic-0 . 36 . jar SE -threads 7 -phred33 [input read FASTQ file] [output read FASTQ file] ILLUMINACLIP:[illumina adaptors sequence FASTA file]:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:50" . The sequence file for Illumina adaptors included both sequences and reverse-complemented sequences for all of the following Illumina adaptor sequences from manufacturer's instructions: TruSeq Universal Adapter and TruSeq Adapter Index 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 18 , 19 , 20 , 21 , 22 , 23 , 25 , and 27 . We assayed expression levels against our previously published protein-coding gene/transcript set for A . ceylanicum , downloaded from the ParaSite database ( release 6; ftp://ftp . wormbase . org/pub/wormbase/parasite/releases/WBPS6/species/ancylostoma_ceylanicum/PRJNA231479/ancylostoma_ceylanicum . PRJNA231479 . WBPS6 . CDS_transcripts . fa . gz ) [39] . We quantitated gene expression from all of our quality-filtered A . ceylanicum RNA-seq data sets with Salmon 0 . 7 . 2 ( https://github . com/COMBINE-lab/salmon/releases/download/v0 . 7 . 2/Salmon-0 . 7 . 2_linux_x86_64 . tar . gz ) , generating expression values in Transcripts Per Million ( TPM ) and estimating mapped read counts per gene [40] . For Salmon's index program , we used the arguments "—no-version-check index—kmerLen 31—perfectHash—type quasi—sasamp 1"; for Salmon's quant program , we used the arguments "—libType A seqBias gcBias numBootstraps 100—geneMap [transcript-to-gene table]" , with "—unmatedReads" specifying the 50-nt single-end data . For gene annotations , we created new Pfam motif annotations with hmmscan from HMMER version 3 . 1b2 [41] and the Pfam 31 . 0 database , using the arguments "—cut_ga -o /dev/null—tblout [table]" which invoked reliably curated domain-specific thresholds; we also generated new InterPro motif annotations with interproscan . sh from InterProScan 5 . 18–57 . 0 [42] , using the arguments "-dp -hm -iprlookup -goterms" . Both Pfam and InterPro motifs were computed solely for the largest isoform of each gene's predicted protein products ( downloaded from ParaSite release 6; ftp://ftp . ebi . ac . uk/pub/databases/wormbase/parasite/releases/WBPS6/species/ancylostoma_ceylanicum/PRJNA231479/ancylostoma_ceylanicum . PRJNA231479 . WBPS6 . protein . fa . gz ) ; these largest isoforms were extracted with get_largest_isoforms . pl using the argument "-t parasite" . All other gene annotations were taken from our previous work [35] . The Perl scripts quality_trim_fastq . pl and get_largest_isoforms . pl are available from https://github . com/SchwarzEM/ems_perl . Adult A . ceylanicum hookworms were collected from the small intestine of a day 22 post-inoculation ( PI ) hamster into a 1 . 5 mL microfuge tube , were rinsed three times with Milli-Q water , and then snap-frozen in liquid nitrogen and stored in -80°C . Tissue homogenization was performed in the same 1 . 5 mL microfuge tube on liquid nitrogen using a pre-chilled tapered flat end weighing spatula followed by a pre-chilled micropestle . Total nucleic acid was isolated with Nucleospin RNA kit ( Machery-Nagel ) according to the manufacturer’s instructions , except that on-column DNase treatment was omitted . The total nucleic acid was then treated with RNase-free DNase I ( NEB ) according to the manufacturer’s instructions . RNA was precipitated by addition of 1:10 vol 3 M sodium acetate and 2 . 5 vol ethanol with O/N storage in -20°C . The RNA pellet was washed twice with ethanol and then dissolved in 50 μL Milli-Q water . cDNA was synthesized with qScript cDNA SuperMix ( Quantabio ) according to the manufacturer’s instructions . PCR was performed with Platinum Taq DNA Polymerase High Fidelity ( Invitrogen ) according to the manufacturer’s instructions using the following primers: Aceys0154g3007cdsF and Aceys0154g3007cdsR ( AceyCP1 ) , Aceys0532g3038t1cdsF and Aceys0532g3038t1cdsR ( AceyCPL ) , and Aceys0034g2829t1cdsF and Aceys0034g2829t1cdsR ( AceySKPI3 ) ( Table S1 ) . The PCR products were purified with Monarch PCR and DNA Cleanup Kit ( NEB ) and were sequenced at GENEWIZ . The validated CDS sequences were sent to Genscript for P . pastoris codon-optimized DNA synthesis and the CDSs without native signal peptides ( AceyCP1 , nt 40–1 , 032; AceyCPL , nt 46–1 , 032; and AceySKPI3 , nt 52–240 ) were subcloned into pPICZαA in-frame with yeast α-factor signal . During subcloning polyhistidine tags were added directly 5’ to the CDSs by PCR . The plasmids were linearized with SacI and transformed into P . pastoris X-33 . Single transformed colonies ( confirmed by colony PCR ) were inoculated into 25 mL Buffered Glycerol-complex Medium ( BMGY ) and grown to an OD600 of 3 . 0 . The 25 mL cultures were used to inoculate 0 . 5 L BMGY cultures at an OD600 of 1 . 0 , and these cultures were grown to an OD600 of 3 . 0 . The BMGY cultures were centrifuged and each was resuspended in 2 L of Buffered Methanol-complex Medium ( BMMY ) distributed into four 2 L baffled flasks , and these cultures were grown for 4 days . After 4 days of incubation , the BMMY cultures were centrifuged , and the clarified supernatants were collected . rAceyCP1 , rAceyCPL and rAceySKPI3 were purified from X-33 culture supernatants by immobilized metal affinity chromatography using a Ni resin and column ( GenScript ) . Proteins bound to the resin were washed with Triton X-100 to reduce endotoxin levels to <1 EU/μg . The eluates were buffer exchanged into PBS ( pH 7 . 4 ) by dialysis , and then filter sterilized with 0 . 22 μm Millex-GP Syringe Filters . Endotoxin levels were detected by ToxinSensor Gel Clot Endotoxin Assay Kit ( GenScript ) . Protein concentrations were determined by Bradford assay using BSA as standard ( GenScript ) . For SDS-PAGE , 4 μg of each protein was boiled for ~5 min in Pierce Lane Marker Reducing Sample Buffer ( Thermo Fisher ) and loaded into a 12% Tris-Glycine mini gel . Electrophoresis was run for ~2 hr at 100V in a Mini-PROTEAN Tetra Cell ( Bio-Rad ) . Proteins were stained with Coomassie Blue , and the gels were imaged with a ChemiDoc XRS+ System with Image Lab Software ( Bio-Rad ) . Molecular weights were estimated in Image Lab from the SDS-PAGE gels . For Western blots , immediately after electrophoresis proteins were transferred to PVDF membranes using a Trans-Blot Turbo Transfer System with RTA Mini LF PVDF Transfer Kit ( Bio-Rad ) . The membranes were blocked for 1 . 5 hr in blocking buffer ( 3% non-fat dry milk prepared in PBST ) . The blocked membranes were washed for 5 min twice with PBST , and then incubated for 1 . 5 hr in 6x-His Tag Monoclonal Antibody ( HIS . H8 ) ( Invitrogen ) diluted 1:2 , 000 in blocking buffer . The membranes were then washed for 5 min 3 times with PBST , and then incubated for 1 hr in Goat anti-Mouse IgG ( H+L ) Secondary Antibody , HRP ( Invitrogen ) diluted 1:3 , 000 in blocking buffer . The membranes were washed for 5 min 3 times with PBST , and then incubated in the dark for 5 min in SuperSignal West Pico PLUS Chemiluminescent Substrate . The membranes were imaged with a ChemiDoc XRS+ System with Image Lab Software ( Bio-Rad ) . Signal accumulation mode was used first to find the optimum exposure time ( 60 sec ) , and then the membranes were washed for 5 min with PBST and then incubated again in substrate . Finally , the membranes were imaged manually with the optimum exposure times . Vaccines were prepared fresh for each immunization on ice in a total volume of 1 . 8 mL ( rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel , rAceyCP1/Alhydrogel trial 1 ) or 2 . 6 mL ( rAceyCP1/Alhydrogel trial 2 ) in sterile 5 mL microfuge tubes . Two hundred twenty-five μg ( rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel , rAceyCP1/Alhydrogel trial 1 ) or 325 μg ( rAceyCP1/Alhydrogel trial 2 ) of immunogen was diluted up to 1 . 575 mL ( rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel , rAceyCP1/Alhydrogel trial 1 ) or 2 . 275 mL ( rAceyCP1/Alhydrogel trial 2 ) in PBS ( pH 7 . 4 ) , and 225 μL ( ( rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel , rAceyCP1/Alhydrogel trial 1 ) or 325 μL ( rAceyCP1/Alhydrogel trial 2 ) of Alhydrogel ( InvivoGen ) was added . Immunogens were adsorbed to Alhydrogel according to the manufacturer’s instructions . For Alhydrogel control , 225 μL ( rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel , rAceyCP1/Alhydrogel trial 1 ) or 325 μL ( rAceyCP1/Alhydrogel trial 2 ) of Alhydrogel was added to 1 . 575 mL ( rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel , rAceyCP1/Alhydrogel trial 1 ) or 2 . 275 mL ( rAceyCP1/Alhydrogel trial 2 ) of PBS ( pH 7 . 4 ) . Final doses were 25 μg of protein and 250 μg of aluminum content ( per dose ) . Hamsters were injected subcutaneously ( SC ) with insulin syringes ( BD ) three times with two-week intervals in the scruff of the neck with 200 μL of vaccine . One week after the final immunization , hamsters were separated into individual cages , and the individual cages were randomly arranged on the shelves . Twelve days after the final immunization , peripheral blood was collected by saphenous venipuncture using PrecisionGlide 20 G x 1” hypodermic needles ( BD ) and SAFE-T-FILL Capillary Blood Collection Tubes–Serum ( RAM Scientific ) . Blood was allowed to clot for at least 30 min at room temperature before centrifugation . The collected serum was stored in -20°C until ELISA . Thirteen days after the final immunization , hamsters were weighed and blood was collected again , but into SAFE-T-FILL Capillary Blood Collection Tubes–EDTA ( RAM Scientific ) . Blood hemoglobin concentrations ( g/dL ) were measured with a STAT-Site M Hgb Hemoglobin Analyzer and Test Cards ( Stanbio ) . Exactly two weeks after the final immunization hamsters were inoculated with ~150 A . ceylanicum L3i by oral gavage . L3i were obtained by coproculture of feces collected from infected hamsters , and had been stored in the dark at room temperature for <2 weeks in BU buffer ( 50 mM Na2HPO4 , 22 mM KH2PO4 , 70 mM NaCl , pH 6 . 8 ) plus PSF ( 100 U/mL penicillin; 100 μg/mL streptomycin; 0 . 25 μg/mL amphotericin B ) before inoculations . This A . ceylanicum line was originally obtained from Dr . John Hawdon at George Washington University . On day 20 PI , hamsters were weighed and blood hemoglobin concentrations measured as before . On day 22 PI , hamsters were placed on fecal collection wires overnight . Two layers of moistened paper towels were placed in the bottoms of the cages underneath the fecal collection wires . On day 23 PI , hamsters were euthanized by CO2 overdose and cervical dislocation ( according to IACUC protocol ) . Small intestines were removed , longitudinally sectioned , and incubated in Hank’s Balanced Salt Solution ( HBSS; Thermo Fisher ) for 45 min at 37°C , 5% CO2 . Hookworm burdens were counted under a stereomicroscope . Fecal pellets were collected from the cage of each hamster , and FECs were measured using a McMaster chamber ( Hausser Scientific ) . Immunogens were coated overnight at 4°C onto Nunc MaxiSorp flat-bottom 96-well plates at 5 μg/ml in carbonate/bicarbonate ( 100 mM ) , pH 9 . 6 coating buffer ( 100 μL/well ) . Wells were washed three times with PBST ( 200 μL/well ) , and then blocked for 1 . 5 hr in blocking buffer ( 5% non-fat dry milk in PBST ) ( 200 μL/well ) at room temperature . Wells were washed two times with PBST ( 200 μL/well ) , and then hamster sera serially diluted in blocking buffer was incubated ( 100 μL/well ) for 1 . 5 hr at room temperature . Wells were washed three times with PBST ( 200 μL/well ) , and then Peroxidase AffiniPure Goat Anti-Syrian Hamster IgG ( H+L ) ( Jackson ImmunoResearch ) was diluted 1:5 , 000 in blocking buffer and incubated in the wells ( 100 μL/well ) for 1 . 5 hr at room temperature . Wells were washed three times with PBST ( 200 μL/well ) , and 100 μL of 1-Step Ultra TMB-ELISA Substrate Solution ( Thermo Fisher ) was incubated in the wells for 30 min . Then 100 μL of sulfuric acid ( 2 M ) was added to the wells , and A450 was measured with a Tecan Safire plate reader . In multiple pilot experiments , we tested Rabbit anti-Syrian hamster IgM-HRP ( Rockland ) , Goat anti-Mouse IgA-HRP ( Thermo Fisher; [43] ) and Goat anti-Mouse IgE-HRP ( Thermo Fisher ) in ELISAs to soluble A . ceylanicum hookworm extract ( HEX; [43] ) using serum from infected and drug-cleared hamsters . Anti-Syrian hamster IgM-HRP gave extremely high background to HEX with serum from uninfected hamsters . Neither anti-Mouse IgA-HRP or anti-Mouse IgE-HRP reacted to HEX with serum from infected hamsters , while Goat anti-Syrian hamster IgG ( mentioned above ) reacted strongly to HEX only with serum from infected hamsters . Thus , only Goat anti-Syrian hamster IgG was useful for serum ELISAs . A . ceylanicum adult hookworms were collected from the small intestines of initially naïve hamsters on day 17 PI . Small intestines were longitudinally sectioned and incubated for 2 hr in HBSS pre-warmed at 37°C in a mini Baermann Funnel apparatus . Every 20–30 min the small intestines were moved around with forceps . The HBSS containing motile hookworms that had migrated through the wire mesh and settled at the bottom of the funnel was poured into a petri dish , and the motile hookworms were hand-picked with a worm picker into Milli-Q water , and rinsed three times . Hookworms were hand-picked into individual wells of a 96-well plate ( two hookworms per well ) containing 50 μL of HCM with 50% heat-inactivated fetal bovine/calf serum [44] replaced by hamster serum ( mHCM: 49 . 5% RPMI 1640 Medium containing L-glutamine without Phenol Red Indicator [Thermo Fisher]; 49 . 5% hamster serum; 1% 100X PSG [100 U/mL penicillin; 100 μg/mL streptomycin; 0 . 292 mg/mL L-glutamine; Thermo Fisher] ) . Each hamster serum group included three wells ( two hookworms/well for a total of six hookworm adults scored per group ) , and the average motility for the three wells was calculated using a standard 3–0 motility index assay ( 3 = highly motile; 2 = less motile; 1 = motile only when stimulated by touch; 0 = immotile ) [45–47] . Motility was monitored for 76 hr ( once per day ) . To address any concerns about subjectivity , each replicate well for each condition was randomized in the setup so that the investigator was blinded as to which well contained which condition during the scoring process . Another set of hamsters was vaccinated with rAceyCP1/Alhydrogel and Alhydrogel control exactly as before in the vaccine trials . Serum IgG responses were evaluated by ELISA as before using A450 readings at 1:100 serum dilutions . Two weeks after the final immunization hamsters were euthanized as before . Spleens were removed with ethanol-sterilized forceps and transferred to 5 mL DMEM-10 cell culture medium ( 89 . 5% DMEM [Dulbecco's Modification of Eagle's Medium; Mediatech]; 9 . 5% fetal calf serum [Thermo Fisher]; 1% 100X PSG; sterilized with 0 . 22 μm filter; stored in 4°C ) in 60-mm petri dish on ice . Each spleen was cut into three pieces with ethanol-sterilized surgical scissors . Each spleen piece was smashed individually between two frosted microscope slides and rinsed back into the petri dish , removing any remaining solid tissue left on the slide with forceps . Each spleen material in DMEM-10 was passed through syringe needle series of 18 , 22 , and 26 G ( BD ) to prepare the splenocyte suspension that was collected into a 15-mL conical centrifuge tube pre-chilled on ice . Splenocyte suspensions were centrifuged at 1 , 500 rpm for 5 min at 4°C . The supernatant was discarded and cell pellet resuspended in 1 mL ACK Lysing Buffer ( Thermo Fisher ) and incubated for 10 min at room temperature . Immediately after , 9 mL of DMEM-10 was added and then centrifuged as before . Each splenocyte suspension was resuspended in 2 mL DMEM-10 , and 1 mL was transferred into each of 2 different wells of a 48-well plate . For each splenocyte suspension , 25 μg of rAceyCP1 was added to one well and PBS ( pH 7 . 4 ) to the other well . Splenocytes were stimulated for 48 hr at 37°C , 5% CO2 . Stimulated splenocytes were transferred to 1 . 5 mL microfuge tubes and centrifuged as before . Supernatants were discarded , cell pellets were resuspended in lysis buffer and vortexed , and RNA was isolated as before . Either 10 μg ( γ-actin , IFN-γ , IL17A , IL10 , and TGF-β ) or 50 μg ( IL4 , IL5 , IL13 , and IL21 ) of RNA ( as determined in primer efficiency tests ) was used as template for qRT-PCR using qScript One-Step SYBR Green qRT-PCR Kit ( Quantabio ) according to the manufacturer’s instructions . Protocol: 49°C for 10 min , 95°C for 5 min , 35 cycles of 95°C for 15 sec and 60°C for 45 sec . Amplification specificities were verified by melting curve analysis . Melting curve protocol: 95°C for 1 min , 55°C for 10 sec and a slow temperature ramp from 55 to 95°C . The primer sets ( Table S1; [48] ) were pre-validated with mean ± S . D . amplification factors of 2 . 0 ± 0 . 3 in primer efficiency tests using the same sample type . qRT-PCR was performed on an Eppendorf Mastercycler RealPlex2 . Four technical replicates were included for every primer set on each RNA sample . Minus reverse transcriptase reactions were also run on every RNA sample for every primer set , and this indicated that the RNA samples had undetectable levels of gDNA . γ-actin was used as reference . We had hoped to look at even more cytokines by qRT-PCR , but we did not have RNA for all biological replicates . Data were analyzed using the 2-ΔΔCT method [49] . All data analyses were plotted in Prism 7 ( GraphPad Software ) . For serum IgG endpoint titers , plotted are the inverses of the final serum dilutions for each serum sample from each hamster within each vaccine group that gave an A450 reading that is >3 S . D . from the mean A450 reading for naïve hamster sera ( n = 8 ) at the same dilution . For hookworm burden , mean indicates the mean hookworm burden amongst all hamsters in each vaccine group . For FECs ( eggs/g ) , mean indicates the egg count per group from all cages in the group at a given time point . For Δweight and Δhemoglobin , mean indicates the mean Δweight and Δhemoglobin amongst all hamsters in each vaccine group . For motility index , mean indicates the mean motility amongst all wells ( 2 hookworms/well , 3 wells/group ) in the group at a given time point . For serum IgG A450 , plotted are the raw A450 readings for each serum sample from each hamster per vaccine group at 1:100 dilution with average Alhydrogel background subtracted . For relative fold change in mRNA , mean indicates the mean relative fold change in mRNA in splenocytes amongst all hamsters in each vaccine group . Statistical 2-sample comparisons between experimental and control vaccine groups were carried out by one-tailed Mann-Whitney , with the hypothesis that successful vaccination will result in decreases in infection parameters ( worm burdens , fecal egg counts ) and improvements in measurements of sequelae ( weight , hemoglobin ) . Nearly identical statistical results were obtained using student’s t test . All comparisons between multiple experimental groups with control ( i . e . , Alhydrogel alone ) group were carried out using one-way ANOVA with Dunnett’s post-hoc test , comparing each group to control . Significant correlations in linear regression analyses ( i . e . , if slopes are significantly non-zero ) were determined by F test .
Transcriptomic analyses revealed that the hookworm A . ceylanicum upregulates proteases and protease inhibitors during infection , which are likely to be important for successful parasitism [35] . Proteases are thought to help hookworms dig through their host's tissues , destroy proteins needed for the host's immune response , and digest proteins in the host's blood [50] . Protease inhibitors are thought to block host proteases needed for the immune response as well [51] . These proteases and protease inhibitors therefore defined an initial set of vaccine candidates and included putatively secreted cathepsin B-like proteases ( CPs ) , which have homologs in other strongylids , and a previously undescribed family of putatively secreted small Kunitz-type protease inhibitors ( SKPIs ) that are strongly upregulated in adult hookworms [35] . We successfully cloned full-length cDNAs for A . ceylanicum CP1 ( rAceyCP1; AceyCP1 [genomic name , Acey_s0154 . g3007] ) , AceyCPL ( rAceyCPL [Acey_s0532 . g3038] ) , and AceySKPI3 ( rAceySKPI3 [Acey_s0034 . g2829] ) ( Fig 1A ) . We expressed the proteins in Pichia pastoris and purified them using standard protocols ( Fig 1B–1E; see Methods for details ) . P . pastoris expression was used for previous hamster-hookworm vaccine trials [52 , 53] . Since the hookworm intestine is a key target of protective vaccine antigens , we examined expression of these genes in published male hookworm intestinal transcriptomic data ( Table S2 ) [38] . AceyCP1 is very strongly expressed in the male intestine at 11 , 600 transcripts per million ( TPM ) , making it 1 . 2% of all transcripts in that tissue . AceyCPL and AceySKPI3 are present but considerably more weakly expressed in the male hookworm intestine at 52 TPM and 1 . 7 TPM respectively . Neither AceyCP1 nor AceyCPL are expressed in L3i ( 0 . 15 and 0 TPM respectively ) . AceySKPI3 has modest expression in L3i ( 14 TPM ) , but this is less than or comparable to expression levels for the A . ceylanicum homologs of APR1 and GST1 ( AceyAPR1 [Acey_s0242 . g3404] , 744 TPM; AceyGST1 [Acey_s0110 . g143] , 44 TPM ) , which in N . americanus are considered acceptable vaccine candidates . By SDS-PAGE and Western blot , three bands were observed for rAceyCP1 at ~42 , ~44 and ~46 kDa ( Fig 1B and 1C ) , two bands were observed for rAceyCPL at ~40 and ~42 kDa ( Fig 1B and 1C ) , and two bands were observed for rAceySKPI3 at ~12 and ~19 kDa ( mostly single band at ~12 kDa; Fig 1D and 1E ) . AceyCP1 and AceyCPL are proenzymes containing N-terminal pro-regions and a CP domain ( Fig 1A ) . Two N-glycosylation sites were identified in AceyCP1 with NetNGlyc 4 . 0 ( http://www . cbs . dtu . dk/services/NetNGlyc ) [54] , both of which are located in its CP domain near putative active site residues ( Fig 1A ) . One N-glycosylation site was identified in AceyCPL within its pro-region ( Fig 1A ) . These predicted N-glycosylation sites match up with the observed banding patterns ( Fig 1B and 1C ) . NetNGlyc 4 . 0 did not identify any potential N-glycosylation sites in AceySKPI3 ( Fig 1A ) . Treatment of rAceyCP1 and rAceyCPL with endoglycosidase ( Endo ) H resulted in shifts to a single ~42-kDa band for rAceyCP1 and a single ~40-kDa band for rAceyCPL ( Fig 1B and 1C ) . According to these results , rAceyCP1 consists of unglycosylated , monoglycosylated , and biglycosylated forms , while rAceyCPL consists of unglycoslyated and monoglycosylated forms . The weak ~19-kDa upper band in rAceySKPI3 could be a form of dimer or a monomer with other post-translational modification ( s ) . Densitometric analysis determined all three recombinant proteins to be >95% pure . rAceyCPL was adsorbed to Alhydrogel and injected SC into Syrian hamsters three times at two-week intervals ( Fig 2 , timeline ) . Twelve days after the final immunization ( Fig 2 ) , rAceyCPL serum IgG titer was determined by indirect enzyme-linked immunosorbent assay ( ELISA ) . All rAceyCPL/Alhydrogel vaccinates ( 8/8 ) had serum IgG titers above background in Alhydrogel controls ranging from 100 , 000 to ≥500 , 000 ( Fig 3A ) . Hamsters were infected with ~150 A . ceylanicum third stage infective larvae ( L3i ) exactly two weeks after the final immunization ( Fig 2 ) . At day 23 post-inoculation ( PI ) , the number of hookworms in the small intestines ( hookworm burden ) were counted at necropsy , and feces were collected for fecal egg counts ( FECs; hookworm eggs shed per g of feces ) . Mean hookworm burden and FEC were statistically similar between rAceyCPL/Alhydrogel vaccinates and Alhydrogel controls ( Fig 3B and 3C ) , indicating that parasitism was unaffected by vaccination . As markers for the clinical pathology caused by hookworm infection , changes in weight and hemoglobin ( Δweight and Δhemoglobin ) were evaluated ( on day 20 post-inoculation ( PI ) prior to necropsy; Fig 2 ) . Consistent with hookworm burden and FEC , mean Δweight and Δhemoglobin were statistically similar between rAceyCPL/Alhydrogel vaccinates and Alhydrogel controls ( Fig 3D and 3E ) , indicating that clinical pathology was also unaffected by vaccination . All rAceySKPI3/Alhydrogel vaccinates ( 8/8 ) had serum IgG titers above background in Alhydrogel controls ranging from 100 , 000 to , remarkably , ≥10 , 000 , 000 ( Fig 4A ) . However , as with rAceyCPL/Alhydrogel , rAceySKPI3/Alhydrogel vaccination did not significantly alter hookworm parasitism ( Fig 4B and 4C ) , or clinical pathology ( Fig 4D and 4E ) compared to Alhydrogel controls . Thus , these results indicate that although both rAceyCPL/Alhydrogel and rAceySKPI3/Alhydrogel induce high serum IgG titers above background with 100% responder rates , these antigens are not protective . Two independent vaccine trials were conducted for rAceyCP1 , the first with eight animals per group and the second with 12 animals per group . In trial 1 , 5/8 vaccinates ( 62 . 5% ) had rAceyCP1-specific serum IgG titers above background in Alhydrogel controls ranging from 4 , 000–10 , 000 ( Fig 5 ) . In trial 2 , 7/12 vaccinates ( 58 . 3% ) had rAceyCP1 serum IgG titers above background in Alhydrogel controls ranging from 2 , 000–20 , 000 ( Fig 5 ) . In trial 1 and trial 2 , hookworm burdens were decreased in rAceyCP1/Alhydrogel vaccinates compared to Alhydrogel control vaccinates by a mean of 31% and 19% respectively , achieving statistical significance in trial 1 ( Fig 6A ) and nearly achieving statistical significance in trial 2 ( Fig 6A ) . Significant decreases in FECs were seen in both trials ( Fig 6B ) . In trial 1 and trial 2 , respectively , FECs were significantly decreased in rAceyCP1/Alhydrogel vaccinates compared to Alhydrogel control vaccinates by a mean of 46% and 26% . Vaccination with rAceyCP1 also generally led to improvements of hookworm sequelae . Δweight was signficantly improved in rAceyCP1/Alhydrogel vaccinates by 7 . 6 g ( Syrian hamsters weigh around 115 g at this age; ~12 weeks ) compared to Alhydrogel controls in trial 1 ( Fig 6C ) , and there was a trend of 2 . 8 g toward improved Δweight in trial 2 ( Fig 6C ) . The Δhemoglobin was statistically similar between rAceyCP1/Alhydrogel vaccinates and Alhydrogel controls in both trials ( Fig 6D ) . In both vaccine trials , we noticed that rAceyCP1 serum IgG responders were qualitatively more protected than nonresponders ( Fig 6; compare blue and orange data points ) . We hypothesized that the responders were protected from infection and sequelae whereas the nonresponders were not . Thus , we reanalyzed them as separate groups , and compared them to Alhydrogel controls . Strikingly , in trial 1 , hookworm burden was dramatically and significantly decreased in responders by 54% compared to Alhydrogel controls ( Fig 7A ) , whereas nonresponders showed no protection compared to Alhydrogel controls ( Fig 7A ) . In trial 2 , hookworm burden was dramatically and significantly decreased in responders by 40% compared to Alhydrogel controls ( Fig 7A ) , whereas nonresponders showed no protection compared to Alhydrogel controls ( Fig 7A ) . Protection from infection based on hookworm burdens was mirrored in FECs . In trial 1 , FECs were dramatically and significantly decreased in responders by 66% compared to Alhydrogel controls ( Fig 7B ) , whereas nonresponders showed no protection compared to Alhydrogel controls ( Fig 7B ) . Accordingly , in trial 2 , FECs were dramatically and significantly decreased in responders by 54% compared to Alhydrogel controls ( Fig 7B ) , whereas nonresponders showed no protection compared to Alhydrogel controls ( Fig 7B ) . Sequelae based on Δweight were also improved . In trial 1 , responders had a dramatically and significantly improved Δweight of 9 . 6 g compared to Alhydrogel controls ( Fig 7C ) , whereas nonresponders showed no improved Δweight compared to Alhydrogel controls ( Fig 7C ) . In trial 2 , responders had a dramatically and significantly improved Δweight of 6 . 0 g compared Alhydrogel controls ( Fig 7C ) , whereas nonresponders showed no improved Δweight compared to Alhydrogel controls ( Fig 7C ) . In trials 1 and 2 , responders had improved Δhemoglobin by 2 . 2 g/dL and 2 . 0 g/dL respectively , although neither was statistically significant ( Fig 7D ) . A complete summary of the efficacy results from rAceyCP1/Alhydrogel vaccine trials 1 and 2 is given in Table 1 , with the two immunogens currently being tested in phase 1 clinical trials included for comparisons . Linear regression analysis was performed on the data to investigate the correlation between rAceyCP1 serum IgG titer and changes in sequelae and parameters of infection . These analyses determined that , in both vaccine trials , rAceyCP1 serum IgG titer highly significantly correlated with all four measures of protection , including Δhemoglobin ( Fig 8 ) . Moreover , serum IgG titers of 10 , 000–20 , 000 in the two trials ( n = 4 ) gave a mean of 64 . 6% decreased hookworm burden and 76 . 9% decreased FECs compared to Alhydrogel controls . These findings indicate that when rAceyCP1 serum IgG titer is sufficiently induced , rAceyCP1/Alhydrogel is a highly protective hookworm vaccine . We hypothesized that serum IgG might be at least partly responsible for vaccine-induced protections in rAceyCP1/Alhydrogel serum IgG responders . As a first test for this hypothesis , we incubated adult A . ceylanicum hookworms obtained from naïve hamsters in a modified hookworm culture medium ( mHCM ) containing 50% antisera from rAceyCP1/Alhydrogel serum IgG responders and nonresponders , as well as rAceyCPL/Alhydrogel vaccinates and Alhydrogel controls . We scored motility over a period of 76 hr; scoring was performed blind relative to treatment condition to prevent bias . Hookworm motility was significantly reduced by 76 hr in rAceyCP1/Alhydrogel serum IgG responder antisera by a mean of 33 . 3% compared to in Alhydrogel control sera ( Fig 9 ) . Conversely , hookworm motilities in serum IgG non-responder antisera and rAceyCPL/Alhydrogel antisera were reduced by just 5 . 6% and were statistically similar to Alhydrogel control sera ( Fig 9 ) , demonstrating specificity of the activity seen above . Thus , the unique toxicity of antisera alone from rAceyCP1/Alhydrogel serum IgG responders is consistent with the observed protections . Vaccination with rAceyCP1/Alhydrogel induced protection that was highly correlated with serum IgG titer , and found to intoxicate/reduce adult hookworm motility in vitro . In order to gain an understanding of the adaptive , cellular immune responses induced in rAceyCP1/Alhydrogel vaccinates , eight hamsters were vaccinated with rAceyCP1/Alhydrogel exactly as in trials 1 and 2 . Peripheral blood was collected one week after the final immunization to measure serum IgG responses . Necropsy was performed exactly two weeks after the final immunization ( i . e . , at the exact time vaccinates were infected with L3i in trials 1 and 2 ) , and splenocyte suspensions were prepared . rAceyCP1/Alhydrogel and Alhydrogel control splenocytes were stimulated ex vivo with rAceyCP1 or PBS . Total RNA was then isolated and used directly as template for quantitative real-time reverse transcription ( qRT ) -PCR using pre-validated primer sets [48] for the following cytokines: IFN-γ ( Th1 ) ; IL4 , IL5 , IL13 ( Th2 ) ; IL17A ( Th17 ) ; IL21 ( Th17/Tfh ) ; IL10 , TGF-β ( Treg ) ( S1 Fig ) . rAceyCP1/Alhydrogel vaccinates ( 4/8 ) gave rAceyCP1 serum IgG responses ( raw A450 readings ) above background in Alhydrogel controls in serum diluted 1:100 ( Fig 10A ) . rAceyCP1-stimulated splenocytes from rAceyCP1/Alhydrogel IgG responders ( n = 4 ) resulted in elevated levels of all three Th2 cytokine mRNAs ( IL4 , IL5 and IL13 ) that were significantly greater than in stimulated splenocytes from Alhydrogel controls ( n = 5; Fig 10B ) . No other cytokines were elevated in splenocytes from rAceyCP1/Alhydrogel serum IgG responders at levels that were significantly greater than in Alhydrogel control , indicating a highly specific , canonical Th2 cytokine recall response . Also , all cytokines in stimulated splenocytes from rAceyCP1/Alhydrogel IgG nonresponders were statistically similar to Alhydrogel control ( Fig 10B ) .
We demonstrate here that the highly expressed , substantially intestine-enriched cathepsin B cysteine protease , AceyCP1 , is a promising protective antigen candidate for vaccination against A . ceylanicum hookworm infection in Syrian hamsters . Two antigens ( AceyCPL , AceySKPI3 ) with much lower transcript levels in the adult A . ceylanicum intestine compared to AceyCP1 were not protective ( Figs 3 and 4 ) . These data are supportive of the value of using intestinal expression to prioritize antigen candidates . Other intestinal CPs were shown to be protective against other blood-feeding gastrointestinal nematodes ( canine hookworm A . caninum , small ruminant parasite H . contortus , N . americanus hookworms ) in animal hosts ( dogs , sheep , hamsters respectively ) [29 , 55–58] , albeit not to the extent that we report here for AceyCP1 ( e . g . , vaccination with N . americanus CP2 gave 29% reduction in worm burdens in hamsters ) . Interestingly , intestinal CPs were shown not to be protective against an STN parasite of cattle that does not ingest blood , Ostertagia ostertagi [59 , 60] , whereas CPs in O . ostertagi ES products were shown to be protective [59 , 61] . Moreover , recently , IgG induced by vaccination with whole worm extracts of Ascaris suum ( another non-blood-feeding STN ) was determined ( in IgG transfer experiments ) to be protective against infection [62] . Furthermore , O . ostertagi intestinal CP components in a larger native protein complex did cross-protect against H . contortus in vaccinated sheep [60] . These important CP enzymes are therefore vulnerable antigens that can be accessed in the gut only by immune factors that are ingested in the blood . Previous studies of intestinal CPs of blood-feeding STNs have localized the CPs within the worm intestinal lumen with serum IgG from vaccinated hosts , and serum IgG has been implicated as the effector component that neutralizes CP digestion of the blood meal [55 , 56 , 58 , 63 , 64] . However , our finding that AceyCPL vaccination gave a strong immune response but no protection confirms that not all cathepsin B cysteine protease antigens are useful for vaccination . rAceyCP1/Alhydrogel serum IgG responders had dramatically decreased hookworm burdens , FECs , and weight losses in two different vaccine trials compared to Alhydrogel controls , whereas nonresponders did not ( Fig 7 , Table 1 ) . rAceyCP1/Alhydrogel is a highly protective vaccine , reducing hookworm burdens and FECs by ~50% and ~60% , respectively , in responders ( Fig 7 , Table 1 ) , and ~65% and ~77% , respectively , when serum IgG titers were ≥10 , 000 . Although there were moderately decreased blood losses in responders compared to Alhydrogel controls ( Fig 7 , Table 1 ) , these results were not significant . On the other hand , blood loss was highly negatively correlated with rAceyCP1 serum IgG titer ( Fig 8 ) . rAceyCP1-induced protection is among the highest seen to date of current hookworm antigens ( Table 1 ) . Consistently , AceyCP1 is expressed 220 and 6 , 800 times more strongly in the A . ceylanicum intestine compared to AceyAPR1 and AceyGST1 ( 11 , 600 TPM versus 52 and 1 . 7 TPM , respectively ) ( Table S2 ) . Furthermore , AceyCP1 is almost completely unexpressed in L3i ( 0 . 15 TPM ) , while AceyAPR1 and AceyGST1 have significant expression levels in L3i ( 744 and 44 TPM , respectively ) . Thus , AceyCP1 is unlikely to be recognized by IgE and to induce urticarial reactions in previously exposed people from hookworm endemic regions [24] , since L3i is the predominant IgE-reactive stage [65] . Although we cannot rule out effects from different adjuvants , hosts , and Ancylostoma species of vaccine studies carried out to date , rAceyCP1 is clearly a positive addition to the hookworm vaccine antigen arsenal . Antisera from rAceyCP1/Alhydrogel responders , but not from nonresponders or rAceyCPL/Alhydrogel vaccinates , reduced adult A . ceylanicum motility in vitro as early as 24 hr , and was significant by 76 hr ( Fig 9 ) . These results are consistent with a model whereby neutralizing serum IgG inhibit AceyCP1 blood digestion within the hookworm gut , thus leading to starvation—comparable to models for other gut protease antigens of blood-feeding STNs [25 , 27 , 56 , 58 , 63] . We observed Th2-specific cytokine recall responses in stimulated splenocytes from rAceyCP1 responders ex vivo that highly correlated with serum IgG titer ( Fig 10 ) . Our findings reinforce the notion that neutralizing IgG in serum ( likely assisted by Th2 cytokines ) is the protective component to gut antigens in blood-feeding STNs . Vaccinated hamsters in trials 1 and 2 gave ~60% responder rates ( Fig 5 ) . Clearly , a major focus in the future will be to improve responder rates , and rAceyCP1 has the potential to teach us more about how to make a better hookworm vaccine . The ~40% nonresponder rate could be due to a number of factors . First , rAceyCP1 was expressed as a secreted , partially glycosylated protein in P . pastoris yeast consisting of unglycosylated , monoglycosylated and biglycosylated forms with relative abundances as follows: biglycosylated > unglycosylated > monoglycosylated ( Fig 1A–1C ) . Glycosylation of rAceyCP1 may play an important role in the IgG nonresponder rate and/or intermediate titers in responders . It has been hypothesized that proper antigen glycosylation plays a critical role in vaccine efficacy of antigens against other gastrointestinal nematode parasites ( reviewed in [66] ) . Expressing rAceyCP1 in other systems with other glycosylation patterns is therefore an important next step . Another potential contributing factor to the ~40% nonresponder rate and intermediate rAceyCP1 serum IgG titers in responders is major histocompatibility complex ( MHC ) class II ( MHC-II ) allelic variation in the outbred hamster colony , which would give variable Th cell responses [67] . Envigo maintains their hamster colony as outbred by non-sib matings and claims that due to high litter average and reproductive vigor , there should be a diversity of MHC alleles segregating in the colony . In conclusion , rAceyCP1/Alhydrogel is a highly protective vaccine against A . ceylanicum hookworm infection in Syrian hamsters when serum IgG is sufficiently induced , which likely requires help from Th2 cytokines . Serum IgG may target AceyCP1 within the gut , thereby neutralizing hookworm digestion of the blood meal . Efforts are underway to improve rAceyCP1 neutralizing IgG responder rate and titers , efforts that will ultimately improve translation to humans . Hookworm is among the most disabling parasitic diseases of the developing world , and our findings provide important information for advancing hookworm vaccine development .
|
Hookworms are voracious , blood-feeding , soil-transmitted nematode parasites . Adult hookworms infect the small intestine , causing iron-deficiency anemia and other complications . Hookworms are among the most disabling parasites of the developing world . Drugs are useful for controlling hookworm disease . However , because people often get reinfected rapidly and parasites can develop drug resistance , a vaccine that provides long-term protection would improve control and help lead to eradication . At present , there is no licensed hookworm vaccine , and progress towards a vaccine has been limited . We identified a cysteine protease in the intestine of the human hookworm Ancylostoma ceylanicum that is among the most strongly expressed genes during blood feeding and that may help digest blood and be essential for hookworm survival . Vaccination of hamsters with this cysteine protease gave high levels of protection when antigen-specific antibodies in the blood were induced . These antigen-specific antibodies also made hookworms less mobile in culture . This cysteine protease is a promising candidate for further investigation as a human hookworm vaccine antigen .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"immune",
"physiology",
"cytokines",
"helminths",
"immunology",
"hookworms",
"vertebrates",
"parasitic",
"diseases",
"animals",
"mammals",
"vaccines",
"developmental",
"biology",
"molecular",
"development",
"infectious",
"disease",
"control",
"hemoglobin",
"digestive",
"system",
"hamsters",
"infectious",
"diseases",
"proteins",
"immune",
"system",
"gastrointestinal",
"tract",
"biochemistry",
"helminth",
"infections",
"rodents",
"eukaryota",
"blood",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"amniotes",
"organisms"
] |
2019
|
A highly expressed intestinal cysteine protease of Ancylostoma ceylanicum protects vaccinated hamsters from hookworm infection
|
Using a computational model of the Caenorhabditis elegans connectome dynamics , we show that proprioceptive feedback is necessary for sustained dynamic responses to external input . This is consistent with the lack of biophysical evidence for a central pattern generator , and recent experimental evidence that proprioception drives locomotion . The low-dimensional functional response of the Caenorhabditis elegans network of neurons to proprioception-like feedback is optimized by input of specific spatial wavelengths which correspond to the spatial scale of real body shape dynamics . Furthermore , we find that the motor subcircuit of the network is responsible for regulating this response , in agreement with experimental expectations . To explore how the connectomic dynamics produces the observed two-mode , oscillatory limit cycle behavior from a static fixed point , we probe the fixed point’s low-dimensional structure using Dynamic Mode Decomposition . This reveals that the nonlinear network dynamics encode six clusters of dynamic modes , with timescales spanning three orders of magnitude . Two of these six dynamic mode clusters correspond to previously-discovered behavioral modes related to locomotion . These dynamic modes and their timescales are encoded by the network’s degree distribution and specific connectivity . This suggests that behavioral dynamics are partially encoded within the connectome itself , the connectivity of which facilitates proprioceptive control .
The exact process through which the nematode Caenorhabditis elegans ( C . elegans ) generates the rhythmic activity necessary for locomotion remains unclear [1] . In many other species , a Central Pattern Generator ( CPG ) is typically the source of rhythmic activity [2–6] . There is insufficient experimental evidence to support the existence of a CPG in the C . elegans neuronal network [7 , 8] . Experimental and computational evidence shows that proprioception within motorneurons plays an important role in driving and modulating forward locomotion [9 , 10] , and it has been hypothesized that this proprioceptive feedback is what ultimately generates rhythmic locomotion [10] , rather than any dedicated circuitry in the neuronal network . Using a computational model for the connectome dynamics of C . elegans [11] , we provide strong theoretical and computational support , through the emerging method of dynamic mode decomposition , for the hypothesis that proprioception within motorneurons does indeed encode and drive rhythmic activity . Critical to assessing how sustained , low-dimensional dynamic activity is generated , is understanding the role the network’s connectivity graph ( its “connectome” ) plays in generating rhythmic motion . The structure of a neuronal network’s connectivity often determines how the network operates as a whole [12 , 13] , encoding key behavioral responses characterized by low-dimensional patterns of activity [14–19] . However , the exact importance of the specific connectivity of a network is unclear , and neuronal network dynamics are often computationally modeled using uniform random networks [20–27] . In C . elegans , however , the structure of the connectome is clearly not random , and it may further play a critical role in helping to generate or facilitate rhythmic responses . This is suggested by the fact that computational models of the connectome can generate motorneuron oscillations related to forward locomotion in response to constant stimuli even without proprioception ( and even when modeling neural dynamics alone , with no coupled muscular , bodily or environmental modeling ) [11] . This suggests that oscillatory , stereotyped responses are , at some level , encoded within the connectome . There is , however , an important caveat to this result: oscillatory output occurs only due to an unrealistic stimulus , consisting of a constant input into the tail-touch mechanoreceptor sensory neuron pair PLM ( i . e . the touch-receptive posterior lateral microtubule cells ) [28] . In the absence of constant stimulus , the neural state will collapse onto a static , stable fixed point , i . e . a state of no movement . This is illustrated in Panel ( A ) of Fig 1 . This is clearly not realistic; the actual worm is not constantly receiving tail-touch stimulus during every moment at which it crawls forward . As illustrated in Panel ( B ) , the system will quickly decay back to static equilibrium after any random stimulus . A more realistic response to an impulse may perhaps look more like Panel ( C ) : if the worm is in a pause state , a momentary stimulus should be capable of driving it into sustained motion . This lack of sustained oscillation can be explained by the model’s lack of feedback , specifically by the lack of stretch-receptive proprioception within B-class motorneurons , which is known to drive and regulate locomotion [9] . This study thus considers the following questions about the model in [11]: Is it consistent with a framework of proprioception-driven locomotion ? If so , do the low-dimensional output patterns encoded by the connectome facilitate proprioceptive control ? In other words , does the system’s equilibrium have a low-dimensional dynamical structure which facilitates responses related to locomotion ? In this manuscript , we demonstrate that proprioceptive feedback is indeed necessary and sufficient for sustained dynamic responses to external input . This is consistent with the lack of biophysical evidence for a central pattern generator driving locomotion , and the evidence that proprioception drives locomotion . Explicitly , we use the spatial location of specific motorneurons to drive them with a sinusoidal traveling wave , approximating strech-receptive proprioception during locomotion . The functional response of the network to this proprioception-like input is optimized by specific spatial wavelengths , specifically optimal locomotion responses are driven by input with spatial scales consistent with C . elegans body shape dynamics , i . e . eigenworm-like structures [29] . We then repeat this investigation for perturbed networks , including a modification in which all but the experimentally-characterized locomotion subcircuit is ablated . This reveals that the motor subcircuit alone generates a functional response nearly identical to that of the full connectome . However , we find that the locomotory motorneurons are not by themselves sufficient , and that locomotory interneurons are crucial to regulating the response , even though they are not stimulated directly . By applying Dynamic Mode Decomposition to the network data , we discover that the dynamics encode six clusters of dynamic modes with timescales spanning three orders of magnitude . Two of these six dynamic mode clusters correspond to previously-discovered behavioral modes related to locomotion . The dynamic modes and timescales are encoded by the network’s degree distribution and specific connectivity . This suggests that behavioral dynamics are partially encoded within the connectome itself , the connectivity of which facilitates proprioceptive control . Thus our results suggest a framework in which the neural network is not the source of spontaneous oscillation , but rather is structured to facilitate specific proprioception-driven oscillation responses . More broadly , our application of Dynamic Mode Decomposition to network dynamics demonstrates its utility at discovering , from activity data alone , the responses which a network may be encoded to promote or inhibit .
Given the lack of evidence for a CPG within the network , it is interesting that the system is able to generate oscillation in response to a non-oscillatory input , and that this oscillation appears related to locomotion [11] . However , it is clearly unrealistic that such oscillation would require a constant , explicit external input , and would otherwise collapse to a fixed point ( i . e . a static neural pattern ) . The dynamical structure of this fixed point , from which we wish to drive the system into sustained oscillatory motion , can be investigated through impulse-response experiments . In each of 100 separate trials , we model the dynamics of the full somatic nervous system of 279 neurons ( where there are 302 neurons total , 282 within the somatic nervous system , and 279 of those which make synaptic connections [30] ) . We perturbed the system from equilibrium with a short stimulus distributed randomly across all 279 neurons . The system was then allowed to freely decay back to the fixed point , and the decaying neuron voltages were recorded ( providing data as shown in Fig 2 ) . We observed that , in all trials , the system decayed back to the same fixed point regardless of input stimulus . We find that these dynamics are well-described by a few modes ( i . e . specific spatial patterns of neural activation ) , each of which decay exponentially bringing the system back to the fixed point . Applying Dynamic Mode Decomposition to the data gives us both these spatial modes and their decay time constants . Interestingly , we find the following: ( 1 ) in all trials the dynamics are well-described by only six modes , ( 2 ) DMD gives approximately the same six modes regardless of the random stimulus direction , and ( 3 ) the time constants of the modes are well-separated and span three orders of magnitude . Examples of DMD modes and the spatial information which they contain are shown in Fig 2 . These modes can be interpreted as the components of a low-dimensional manifold to which the dynamics are constrained around the fixed point . In other words , an arbitrary stimulus into all 279 neurons can effectively only excite some combination of these six neural patterns . This is what we mean by the fixed point having “low-dimensional structure” . How does this low-dimensional structure relate to the previously-observed , locomotion-like oscillatory response ? To answer this , we note that the PLM response in [11] is characterized by three modes ( the “PLM modes” , which define the spatial activity patterns implicated in this response ) : ( 1 , 2 ) the two modes defining the plane in which the limit cycle proceeds ( the “PLM plane” ) , which we call the “plane modes” , and ( 3 ) the displacement between the equilibrium fixed point and the center of the limit cycle , which we call the “displacement mode” . These modes and their relation to the fixed point and limit cycle are depicted in the “Phase-Plane Dynamics” illustration of Fig 3 . We investigate the biological meaning of our dynamic modes by calculating their projections onto the PLM modes . This reveals that two of the six dynamic modes correspond to previously-discovered PLM modes . These projection values quantify the similarity between the spatial patterns of the previously-discovered , physiologically meaningful PLM modes and the DMD modes which we inferred from our random impulse trials . Fig 4 shows the magnitude of each dynamic mode’s projection onto the displacement mode and PLM plane . Corresponding numerical values are given in Tables 1 , 2 and 3 . We compare these against the projections of 1 , 000 random modes . Both the displacement mode and the PLM Plane have a single dynamic mode projecting strongly onto them . Therefore the low-dimensional structure of the fixed point facilitates responses in both the displacement mode and PLM Plane directions , with a highly distinct timescale for each response . Mode 4 , which projects strongly onto the displacement mode , is particularly interesting . It has the most consistent timescale between trials ( see Table 1 ) . Additionally , all other dynamic modes have a particularly low projection onto the displacement mode ( i . e . a significantly lower median projection than random modes ) . This suggests that the low-dimensional structure of the fixed point facilitates responses in that direction with a particularly consistent timescale . Thus periodic perturbations of the correct timescale could perturb the system off of the fixed point in the direction of the limit cycle; we discuss these implications further in the Discussion . The role of the connectome , the experimentally validated network connectivity , was investigated by repeating the perturbation experiments but with randomly changed network connectivities . We considered the following variations: ( A ) a network with the same degree distribution , with node degrees and connections randomly assigned; ( B ) random connectivity with the same total number of edges . Results from these cases are summarized in Fig 5 . In ( A ) , where the degree distribution is maintained , there are still six modes . However , timescales vary somewhat from the original , and the projections are completely changed . In ( B ) , which has a different degree distribution , the number of modes and all of their properties are qualitatively different . This establishes that the dynamic modes and timescales are encoded by both the network’s degree distribution and specific connectivity . We have established that random impulses can drive the system in the direction of the displacement mode . However , given the apparent global stability of the fixed point , an additional mechanism is required for sustained dynamic responses to external input . Proprioception may allow initial perturbations to grow into the desired limit cycle associated , for instance , with forward motion . We thus investigated the following question: could stretch-receptive proprioceptive feedback within B-class motorneurons give rise to motorneuron oscillations which are qualitatively similar to PLM-driven oscillations ? Since locomotion consists approximately of sinusoidal bends propagating along the body [31 , 32] , it suffices to drive B-class motorneurons with sinusoidal inputs , as determined by their location along the axis of the body and their position on the dorsal/ventral side ( noting that the C . elegans lays on its side while it crawls [32] ) . Fig 6 shows the motorneuron dynamics resulting from different sinusoidal inputs into B-class motorneurons . For certain spatial wavelengths , a limit cycle does occur which is qualitatively similar on the PLM plane . Note that there is a smooth transition between the rows of Fig 6 , and that the middle row is the most similar to the PLM-driven cycle ( as quantified by the Procrustes distance , a measure of shape similarity used more extensively later in the text ) . This middle row corresponds to a spatial driving wavenumber of k = 0 . 886 or , equivalently , a spatial driving wavelength ( per unit body-length ) of λ/L = 1/0 . 886 = 1 . 13 . This value lies well within the range of body-shape wavelengths which , depending on the resistance of the environment , are seen to fall into the range λ/L ∈ ( 0 . 5 , 1 . 75 ) ( specifically , see Fig 1 ( e ) of [33] ) . Note that the variation within this range depends directly upon the environmental resistance , which is not included within our model . Temporal frequency did not have an effect on the shape of the limit cycle , consistent with the experimental observations that the spatial wavelength of C . elegans locomotion does not depend on temporal frequency [32] . As discussed further in the Discussion , this suggests that our model , if integrated with a mechanical body model in future work , could be made consistent with a system of feedback-driven oscillations . An advantage of a full-connectome modeling approach is that it readily enables simulated ablation experiments , in which we may simulate the network with an arbitrary subset of neurons removed . Physiologically , we expect that the response to driving the B-class motorneurons should depend not only upon the neurons which are being driven directly , but also upon associated neurons within the subcircuit of the connectome regulating locomotion , as described in [28 , 34–37] and pictured in Fig 7 ( A ) . We therefore demonstrate that the experimentally-characterized locomotion subcircuit is , by itself , sufficient to reproduce these results . Furthermore , we elucidate the roles of each of its components . We therefore investigate the following questions: is the experimentally-characterized locomotion subcircuit by itself sufficient to reproduce these results ? If so , what is the relative contribution of its different components ? Fig 7 ( B ) shows the PLM-Plane cycle in response to sinusoidal driving of B-class motorneurons , as in Fig 6 ( using a driving wavenumber k = 0 . 886 ) . In addition to calculating this response using the full Connectome , we repeat this simulation with various portions of the network ablated: ( 1 ) with all neurons ablated except for the locomotion subcircuit; ( 2 ) keeping only the locomotory inter-and motorneurons; ( 3 ) keeping only the locomotory motorneurons . We observe that the locomotion subcircuit alone , with the rest of the connectome ablated , reproduces a nearly identical cycle shape . Similarly , when only locomotion inter- and motorneurons are included , the cycle is minimally distorted . However , ablating the locomotory interneurons causes considerable distortion of the cycle , despite the fact that these neurons are not driven directly . Thus we find that these interneurons are crucial in regulating the driven response , which is consistent with evidence for the role of these interneurons in locomotion [28 , 36] . We can quantify the degree to which the cycle is distorted by taking the full connectome’s cycle and the ablated connectome’s cycle and computing their Procrustes distance ( a measure from statistical shape analysis which increases as the cycle shapes become increasingly dissimilar ) . The Procrustes Distance ( P . D . ) of each cycle appears in Fig 7 ( B ) , below each distorted cycle . We further used this to calculate the amount of cycle distortion when each individual component of the locomotory circuit was ablated from the full connectome , allowing for the assessment in each neuron’s relative importance in regulating the response . The individual ablations leading to the highest degree of distortion are included in Fig 7 ( C ) . Notably , this identifies neurons known to be crucial for the worm’s locomotion ability: experimentally ablating AVB and DVA , for example , are each known to cause significantly distorted forward locomotion [28 , 36] .
In this manuscript , we have introduced the Dynamic Mode Decomposition as a diagnostic tool to characterize impulse-response experiments on a nonlinear networked system . This revealed that the network is structured to generate a low-dimensional response at distinct timescales ranging over several orders of magnitude , and that two of these dynamic modes are related to the previously-characterized “forward motion” response to PLM-stimulation . It is possible that proprioceptive feedback could sustain a limit cycle but not be sufficient to bring the system to said limit cycle from the equilibrium fixed point . In other words , the limit cycle would need to be “jumpstarted” , with a separate mechanism transporting the system from the fixed point near to the cycle . Were this the case , it would suggest a physiological purpose for the low-dimensional fixed-point structure which we detect: stimuli of the correct timescales could selectively perturb the system towards the limit cycle , to a point from which the proprioceptive feedback could be effective . In this view , it is interesting and suggestive that Mode 4 , associated with the Displacement Mode , has the most tightly-constrained timescales of all the modes . Repeating this analysis for different connectivities suggested that these dynamic modes and timescales are encoded by both the network’s degree distribution and specific connectivity . A random graph , with the same number of nodes and connections but a different degree distribution , leads to a completely different number of modes . This suggests that the number of dynamical timescales is encoded by the degree distribution , as six timescales are recovered for any network with the same degree distribution . However , the specific timescale values and the neuronal makeup of these modes is not preserved . The degree to which each mode projects onto our biophysiologically-relevant directions , and with what specific dynamical timescale , depends on the specific wiring of the connectome . Thus behavioral dynamics are partially encoded within the connectome itself , the connectivity of which facilitates proprioceptive control . Said another way , the stereotyped worm connectome seems to be optimized for its behavioral repertoire . The usefulness of these insights as they apply to the actual system , however , depend on the model’s compatibility with a framework of proprioception-generated oscillation . Thus we further show that sinusoidal input into the putatively proprioceptive B-class motorneurons does , indeed , drive a limit cycle at certain spatial wavelengths , consistent with the spatial wavelengths seen experimentally . Given that the worm crawls with a sinusoidal body shape [31 , 32] , this suggests that motorneuron proprioception could indeed drive the limit cycle , which in turn could drive sinusoidal movement . A proprioceptive mechanism such as this is necessary for sustained dynamic responses to external input . Furthermore , we showed that the motor subcircuit alone is capable of sustaining these results , and that this circuit’s interneurons are crucial to regulating the response despite not being driven directly . Indeed , we found that the interneurons which were most important to regulating the response within our simulation were those which have been shown to have this exact role experimentally . Despite this apparent consistency , the development of such a feedback rule remains nontrivial . Without a coupled biomechanical model that includes muscle activation , any feedback rule which we might implement on the present model would be no less artificial than our direct sinusoidal stimulus , which is biophysiologically reasonable . However , modeling the worm’s body and environment is ultimately crucial to fully understanding its behavior [7 , 10 , 33 , 38–40] . This study prescribes multiple studies for future computational connectome models which are fully integrated with biomechanical body and environmental models ( as exemplified by projects such as OpenWorm [41] ) . Specifically , it introduces the following questions: ( 1 ) When motorneuron proprioception and other external feedback is turned off within a model , does the system decay into a fixed point ? If so , an identical study can be performed to probe that fixed point’s low-dimensional structure . ( 2 ) Do the dynamic modes relate to the oscillatory dynamics which occur during locomotion ? ( 3 ) If proprioception/feedback is turned back on while the system is in its fixed point , does the system proceed into a spontaneous limit cycle , and if so , how ? Is periodic noise or other stimulation of a specific timescale necessary for such a transition ? More broadly , this work demonstrates the utility of Dynamic Mode Decomposition in relating the specific connectivity of a network to the multi-scale , low-dimensional structure of its dynamical responses . The methods of this manuscript are able to directly relate connectivity to dynamics even for large , nonlinear networked systems . Future work will further investigate this relationship , with implications for the design of nonlinear networks .
Our model for the C . elegans simulates the neuronal dynamics of its full connectome , as obtained from [30] . This network consists of the 279 somatic neurons which make synaptic connections . Between these neurons , there are 6393 synaptic connections and 890 gap junctions , and the connectivity between neurons cannot be considered sparse . Further details on the network’s structural properties are available in [30] , and further information , including about putative functions of individual neurons , is collected within WormAtlas [42] . Experiments show that many neurons in the organism are effectively isopotential , such that membrane voltage is a meaningful state variable [43] . Wicks et al . constructed a single-compartment membrane model for neuron dynamics [44] , which we later adapted to incorporate recent connectomic data [11] . We assume that the membrane voltage dynamics of neuron i is governed by: C i v i ˙ = - G i c ( v i - E c e l l ) - I i G a p ( v ) - I i S y n ( v ) + I i E x t . ( 1 ) The parameter Ci represents the whole-cell membrane capacitance , G i c the membrane leakage conductance and Ecell the leakage potential of neuron i . The external input current is given by I i E x t . Neural interaction via gap junctions and synapses are modeled by the input currents I i G a p ( v ) ( gap ) and I i S y n ( v ) ( synaptic ) . Their equations are given by: I i G a p = ∑ j G i j g ( v i - v j ) ( 2 ) I i S y n = ∑ j G i j s s j ( v i - E j ) ( 3 ) We treat gap junctions between neurons i and j as ohmic resistances with total conductivity G i j g . We assume that I i S y n is also modulated by a synaptic activity variable si , which is governed by s i ˙ = a r ϕ ( v i ; κ , v t h ) ( 1 - s i ) - a d s i . ( 4 ) Here ar and ad correspond to growth and decay time , and ϕ is the sigmoid function ϕ ( vi;κ , vth ) = 1/ ( 1 + exp ( −β ( vi − vth ) ) ) . Simulations were performed in MATLAB via Euler’s method , using timesteps of h = 10−6s . The data was downsampled by recording v ( t ) every Δt = 3 × 10−5s , yielding a data matrix: V = | | | v ( t 1 ) v ( t 2 ) ⋯ v ( t m - 1 ) | | | , ( 5 ) where tk+1 − tk = Δt . The value of Δt was chosen to be sufficiently low so as to not affect the outcome of the analysis . We keep all parameter values from [11] . The number of gap junctions N i j g and number of synapses Nijs are taken from the large component of the full connectome , i . e . the 279 neurons as considered in Varshney , et al . [30] . Each individual synapse and gap junction is assigned an equal conductivity of g = 100pS ( such that G i j g = g · N i j g and G i j s = g · N i j s ) . The values of cell membrane conductance and capacitance are Gc = 10pS and C = 1pF . The synaptic growth and decay constants are kept as ar = 1 s−1 and ad = 5 s−1 . All neurons are modeled as identical except for their connectivity and the assignment of them as excitatory or inhibitory ( where Ej will have one of two values corresponding to these classes ) . For each random perturbation simulation , a random external input IExt was applied to all neurons for a duration of 10−5s , after which the system was allowed to decay . Output was recorded from all neurons after the cessation of input . Each I i E x t was drawn from a Gaussian distribution , after which the total IExt was then normalized to have a fixed total input amplitude of |IExt| = 10mA . This section describes the method of Dynamic Mode Decomposition [45–52] , which we apply to our simulated neural voltage data V . Specifically , we use it to relate the voltages at timestep tk to the following timestep tk+1 as follows: v ( t k + 1 ) ≈ A v ( t k ) , ( 6 ) where A ∈ R n × n is the linear operator which is the best-fit solution for all pairs . Note that this does not imply that the underlying dynamics are linear; DMD is connected to nonlinear dynamical systems through the Koopman operator [50] . We can express this relationship in matrix form by constructing two data matrices X ∈ R n × ( m - 1 ) and X ′ ∈ R n × ( m - 1 ) as follows: X = | | | v ( t 1 ) v ( t 2 ) ⋯ v ( t m - 1 ) | | | , ( 7 ) X ′ = | | | v ( t 2 ) v ( t 3 ) ⋯ v ( t m ) | | | . ( 8 ) This allows us to write Eq ( 6 ) as: X ′ ≈ A X . ( 9 ) The dynamic mode decomposition of the data matrices ( X , X′ ) is given by the leading eigendecomposition of the matrix A , which is defined as follows: A = X ′ X † , ( 10 ) where † denotes the Moore-Penrose pseudoinverse [47] . The pseudoinverse of X can be found by calculating its singular value decomposition , truncated at r singular values: X ≈ U ˜ Σ ˜ V ˜ * . ( 11 ) Here * denotes the complex conjugate transpose , U ˜ ∈ R n × r and V ˜ ∈ R m - 1 × r are matrices with orthonormal columns , and Σ ˜ ∈ R r × r is diagonal . The diagonal entries of Σ are the singular values , and are proportional to the percentage of energy within each mode . We choose the smallest set of r modes which capture 99% of the energy . We can thus approximate the linear operator A as follows: A ≈ A ¯ = X ′ V ˜ Σ ˜ - 1 U ˜ * . ( 12 ) We are interested in the dynamics projected upon the lower-dimensional subspace as defined by the first r columns of U ˜ . Rather than calculating the n × n matrix A ¯ , we project onto the low-dimensional subspace to calculate the r × r reduced order operator A ˜: A ˜ = U ˜ * X ′ V ˜ Σ ˜ - 1 . ( 13 ) The eigendecomposition A ˜ W = W Λ gives the eigenvectors wj and eigenvalues λj of the reduced-order system . The eigenvalues are equal to those of the full-dimensional A ¯ , and the corresponding eigenvectors can be used to exactly calculate the full-dimensional dynamic modes of the system [47] . For λj ≠ 0 , the dynamic mode corresponding to wj is: ϕ j = X ′ V ˜ Σ ˜ - 1 w j . ( 14 ) The DMD modes take the eigenvectors of the reduced-order system and project them back to the full-dimensional space . In our system , this means that a dynamic mode ϕj will be a vector of length 279 , with each element corresponding to the relative activation of a neuron within each mode . Since these dynamic modes correspond to the eigenvectors of the low-dimensional system , the modes give the dynamically-decoupled low-dimensional patterns which will exponentially grow/decay and/or oscillate with timescales given by their respective eigenvalues λj . The state of the system just after perturbation may be written in terms of these modes: v ( t = 0 ) ≈ ∑ j = 1 r c j ϕ j . ( 15 ) After k timesteps Δt = tk+1 − tk , the system will then be within the state: v ( t k ) ≈ ∑ j = 1 r c j λ j k ϕ j . ( 16 ) We can also write the solution for an arbitrary time t as: v ( t k ) ≈ ∑ j = 1 r c j ϕ j exp ( - t / τ j ) . ( 17 ) The continuous decay constant τj can be directly calculated from the DMD eigenvalue as follows: λ j = exp ( - Δ t / τ j ) → τ j = - Δ t ln ( λ j ) . ( 18 ) In general , τj may be complex with any sign . Clearly , Re ( τj ) > 0 will lead to exponential decay , Re ( τj ) < 0 will lead to exponential growth , and Im ( τj ) ≠ 0 will lead to oscillation . For all trials within this manuscript , however , the resulting τj values were seen to be positive and real , due to the dynamics of the dataset being well described by non-oscillatory decay . The properties of the resultant modes are summarized in the boxplots of Figs 4 and 5 . These were generated from MATLAB function boxplot . m . Default settings are used in Fig 4 , and in Fig 5 the settings are changed such that no points are treated are outliers . The PLM modes are calculated by taking the singular value decomposition of the PLM-driven limit cycle , as in [11] ) . The PLM modes include the displacement mode d and plane modes p1 , p2 , as illustrated in Fig 3 . Each of these modes are a vector of length 279 , with each element corresponding to the relative activation of a neuron’s membrane voltage . Each represents a neural pattern which is dynamically active which the system while it is in the PLM-driven limit cycle , which is argued in [11] to represent a neural proxy for forward locomotion . The projection metrics are defined as the projections of each dynamic mode vector ϕi onto the displacement mode vector d and onto the PLM Plane {p1 , p2} . Specifically: Disp . Mode Projection = ϕ i · d ( 19 ) Plane Projection = ( ϕ i · p 1 ) 2 + ( ϕ i · p 2 ) 2 ( 20 ) The random projections in Fig 4 are calculated similarly , but using a randomly-generated mode in place of an actual DMD mode . For each random mode , each element is chosen from a Gaussian distribution and the mode is then normalized . We repeated our Dynamic Mode Decomposition analysis for altered networks with ( A ) the same degree distribution , but altered specific connectivity , and ( B ) random graphs with the same total number of connections . For both cases , we generated 5 distinct altered networks , for which we performed 10 impulse-response trials each . We calculated the DMD results ( decay constants τ and displacement mode/PLM plane projections ) for each set of trials . Thus for both ( A ) and ( B ) we obtained 5 sets of 10 τ/projection values each , each set corresponding to a different altered network . Fig 3 plots the distributions of results for all altered networks of a given type ( i . e . plotting all 50 values for each mode ) . All altered connectomes with the same degree distribution , in all of their random-impulse response trials , yield six dynamic modes , as shown in Fig 3 ( A ) . This is the same number modes as is produced by the standard connectome , and thus the distributions of τ and projection values may be directly compared . We wish to determine if the results which we obtain from the altered connectome are statistically different from those which we obtain from the standard connectome . For each of the 5 altered connectomes , we compare the altered and standard τ/projection distributions using the two-sample Kolmogorov-Smirnov test , with the null hypothesis that they are from the same distribution ( computed in MATLAB using the built-in function kstest2 . m ) . The maximum p-values for each distribution from the set of tests is shown in Table 4 . At a significance level of p = 0 . 05 , we can conclude that altering the specific connectivity alters the following results: the τ values of Modes 2 , 3 , 4 , 5 and 6; the displacement mode projections of Modes 4 , 5 and 6; and the plane mode projections of Modes 3 , 4 and 6 . All of the random graphs , in all of their trials , yielded four dynamic modes , as shown in Fig 3 ( B ) . As is apparent in the Fig , Mode 3 is notable for having very consistent projection values onto each mode . Note , however , that this mode is trivial: it is simply equal for all nodes ( i . e . it is the vector ϕ i = 1 / 279 for each of the i ∈ ( 1 , 279 ) neurons ) . This mode , though trivial , will have a higher projection value than will sparser modes which select the “wrong” neurons . In approximating proprioceptive input , we sinusoidally drove all B-class motorneurons , using an external input of the following form: I i E x t = ± A sin ( ω t - k x ) ( 21 ) Input sign was given based on the dorsal/ventral location of the motorneuron . Input amplitude A affected only the amplitude of the cycle and was set at A = 30 Arb . Units to yield a qualitatively similar cycle amplitude . Temporal frequency ω appeared to affect the response only by changing the cycle period . Spatial wavelength k varied between trials ( as shown in Fig 6 ) . x was assigned to each neuron based on its soma position . Soma position data originates from [30] , and was retrieved from the “Neuronal Wiring” section of WormAtlas [42] . The use of the soma position is a simplification: proprioception in B-class motorneurons is believed to be due to stretch reception within the long axons posterior to the soma [9] . The plane dynamics plotted in Figs 6 and 7 were calculated by taking the projection of the full-dimensional dynamics v ( t ) onto the plane modes p1 and p2 . This gives the cycle dynamics projected into the low-dimensional space , as in [11] . We also calculated the response to sinusoidal driving of B-class motorneurons for networks with various neurons removed . We were particularly interested in the role of the motor subcircuit in regulating the response . As in [34] , we take the following neuron groups as comprising the motor circuit: “Simulated ablation” of a neuron is done similarly to how it was performed in [11] , i . e . by simply removing the connections of selected neurons . In other words , we use the same model , with the connectivity data altered such that: G i j g = G i j c = 0 if i or j ablated ( 22 ) Specifically , we calculate the driven limit cycle response , projected onto the PLM plane , for the following ablation sets: Procrustes Distance ( PD ) measures the dissimilarity between shapes , and we use it to quantify the similarity between the shapes of the limit cycles pre- and post-ablation . We use the function procrustes . m from MATLAB’s Statistics and Machine Learning Toolbox . We collect N data points from each trajectory and annotate their ( x , y ) coordinates in a ( 2 × N ) shape matrix S . The PD between two distinct shapes SA and SB is then given by P D = min b , R , c || S B - b · S A · R + c → || 2 . ( 23 ) In other words , it finds the optimal ( 2D ) rotation matrix R , scaling factor b > 0 , and translation vector c → to minimize the sum of the squares of the distances between all points . Intuitively , it compares the shapes of the cycles while ignoring any translation , rotation , or scaling . Note that trajectories must be pre-processed to extract data points for a single period of the cycle . Cycles are also interpolated using MATLAB’s spline . m function to ensure that they have the same number of data points . Both limit cycles must also be phase-aligned , which we achieve by finding the relative phase that minimizes the Procrustes Distance . This results in a score which increases as the post-ablation cycle becomes increasingly dissimilar in shape to the pre-ablation cycle .
|
The nematode C . elegans lives a complex and rich life despite having only 302 neurons . The full connectivity between these neurons ( its “connectome” ) has been measured , making it an ideal model system for understanding how neural processing generates behavior . However , unlike most animals , it doesn’t appear to have neural circuits dedicated to generating rhythmic motion . Even the simple rhythmic behavior of forward locomotion is believed to be ultimately driven by stretch-receptive proprioception as the worm changes body shape . Computational models of connectome dynamics can capture behavioral responses , making them valuable for studying neural dynamics . We consider a model of the full-connectome dynamics which , without external input , is in a static neural state . We show that proprioception-like input can bring the system into a rhythmic state similar to locomotion . We then show that arbitrary external inputs across all neurons can only excite some combination of six neural patterns , and that the shape and dynamics of these patterns are determined by the connectome . Furthermore , two of these patterns correspond to forward locomotion , such that the connectome facilitates locomotion-like responses . We thus find that the connectome itself is not random , but is ideally tuned for generating behavior .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"nervous",
"system",
"caenorhabditis",
"social",
"sciences",
"neuroscience",
"animals",
"biomechanics",
"biological",
"locomotion",
"motor",
"neurons",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
"brain",
"mapping",
"experimental",
"organism",
"systems",
"interneurons",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"connectomics",
"cellular",
"neuroscience",
"psychology",
"proprioception",
"neuroanatomy",
"cell",
"biology",
"anatomy",
"physiology",
"neurons",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"sensory",
"perception",
"organisms"
] |
2017
|
Spatiotemporal Feedback and Network Structure Drive and Encode Caenorhabditis elegans Locomotion
|
Each Caulobacter cell cycle involves differentiation and an asymmetric cell division driven by a cyclical regulatory circuit comprised of four transcription factors ( TFs ) and a DNA methyltransferase . Using a modified global 5′ RACE protocol , we globally mapped transcription start sites ( TSSs ) at base-pair resolution , measured their transcription levels at multiple times in the cell cycle , and identified their transcription factor binding sites . Out of 2726 TSSs , 586 were shown to be cell cycle-regulated and we identified 529 binding sites for the cell cycle master regulators . Twenty-three percent of the cell cycle-regulated promoters were found to be under the combinatorial control of two or more of the global regulators . Previously unknown features of the core cell cycle circuit were identified , including 107 antisense TSSs which exhibit cell cycle-control , and 241 genes with multiple TSSs whose transcription levels often exhibited different cell cycle timing . Cumulatively , this study uncovered novel new layers of transcriptional regulation mediating the bacterial cell cycle .
The regulation of timing and ordered progression of cell cycle events is central to the survival of any organism and is one of the fundamental processes of life . The gram-negative α-proteobacterium Caulobacter crescentus ( Caulobacter , hereafter ) is an important model organism for the study of the regulation of cell cycle progression and asymmetric cell division , shown in Fig . 1A [1]–[3] . A hallmark of Caulobacter asymmetric cell division is that the daughter stalked cell immediately initiates DNA replication and the daughter swarmer cell has a period of motility before differentiating into a stalked cell and initiating chromosome replication . Control of cell cycle progression and asymmetric division occurs through coordinate regulation of transcription , protein phosphorylation , DNA methylation , protein localization , and protein degradation [1] , [4]–[6] . A cyclical genetic circuit , comprised of five master regulator proteins , including DnaA , GcrA , CtrA , and SciP , and the DNA methyltransferase CcrM , drives the cell cycle [2] , [4] ( see Fig . 2B ) . The circuit regulates the transcription of more than 200 genes controlling sequential polar differentiation events including flagella biosynthesis , pili biosynthesis , chemotaxis complex formation , DNA-replication , and cell division [3] , [7]–[12] . However , the mechanism of cell cycle control for only a subset of these has been described . To decipher the regulatory landscape that guides the cell cycle we need to identify transcription start sites ( TSSs ) , measure their cell cycle stage-specific levels , and define the regulatory motifs within each cell cycle-regulated promoter . Here , using a detailed map of the coding and non-coding features in the genome based upon ribosome profiling [13] , we applied global 5′ RACE to map approximately 75 percent of the Caulobacter TSSs at single base-pair resolution and to measure the abundance of RNAs carrying a 5′ tri-phosphate ( 5′ PPP ) group . This was done at multiple time points during the cell cycle to determine the timing of activation of TSSs . We also identified binding sites of the key cell cycle regulatory transcription factors directly upstream of the TSSs . When multiple TFs were predicted to bind within these TSS-proximate regions , we were able to provide an initial estimate of the combinational control logic . For example , the core cell cycle circuit regulators DnaA and GcrA often regulate gene expression in combination with other transcription factors . For genes controlled by the cell cycle circuit regulator CtrA , the location and presence of full palindromic or half CtrA binding motifs , and co-appearance of SciP binding motifs , dictates the cell cycle timing of their transcriptional regulation . We discovered that 107 antisense TSSs positioned within Coding DNA Sequences ( CDSs ) are temporally regulated and identified 241 genes transcribed from multiple promoters whose activation is independently controlled , yielding different timing of TSS activation . Furthermore , we found internal promoters in operons that were independently regulated to alter the expression profiles of encoded genes . Cumulatively , these observations suggest that the regulation of Caulobacter TSS levels during the cell cycle is much more complex than previously reported and this dataset provides a powerful resource for the elucidation of the cell cycle regulatory circuit .
We used a global 5′ RACE ( rapid amplification of cDNA ends ) method in combination with Illumina high-throughput sequencing to obtain a single-nucleotide resolution map of TSSs and their cell cycle-dependent activation level . Isolated swarmer cells ( 0 minutes ) were grown in M2G minimal media for 140 minutes until cell division ( Fig . 1A ) . We collected cell samples at 8 time points during the cell cycle and carried out total RNA extraction to prepare an Illumina high-throughput sequencing library for each time sample ( S7 Fig . and Materials and Methods ) . At each TSS the 5′ nucleotide contains a 5′ PPP group , whereas products of nuclease cleavage yield either a 5′ mono-phosphate ( 5′ P ) group or a 5′ hydroxyl ( 5′ OH ) group . As many processed RNAs such as mature ribosomal and transfer RNA ( rRNA , tRNA ) have a 5′ P , we prepared two additional libraries from unsynchronized culture in mid-exponential growth in minimal media to selectively distinguish between RNA segments with 5′ PPP ends and 5′ P ends . In one of these two libraries , the RNA was treated with tobacco acid pyrophosphatase ( denoted +TAP ) to hydrolyze the 5′ PPP to 5′ P . The other library ( denoted -TAP ) was prepared without TAP treatment ( S7 Fig . ) . Libraries were ligated with a 5′ sequencing adapter followed by reverse transcription using a random-hexamer primer conjugated with a second Illumina sequencing adapter ( S7 Fig . ) . Since T4 RNA ligase reacts only with a 5′ P , removal of the pyrophosphate group allows for the ligation of the 5′ sequencing adapter [14] . About 180 million 38 bp reads were obtained from the ten sequencing libraries ( 8 TAP treated cell cycle time point libraries and the +TAP and -TAP libraries from an unsynchronized mid-log phase culture ) . The reads were aligned onto the Caulobacter NA1000 genome DNA sequence NC_011916 [15] using Bowtie 0 . 12 . 7 software [16] . Only non-rRNA reads that mapped to a unique genomic location without mismatches were used for our analysis . Sequencing reads for all libraries were normalized to the total number of non-rRNA reads in each library . To identify TSSs , we used 34 biochemically-characterized TSSs as a positive control ( S2 Dataset ) and 24 tRNA 5′ P-sites as a negative control ( n = 24 ) ( S1 Fig . ) . We compared the natural log of the ratio ( θ ) between the number of sequencing reads obtained at the 5′ ends of positive and negative controls in +TAP/-TAP libraries . The θ's obtained for positive and negative controls fall into two separate normal distributions with slight overlap ( Welch two sample t-test , df = 41 . 2 , p-value = 1 . 3 e−11 , S1 Fig . ) . To minimize false-positive TSSs , we set the threshold value at θ>0 . 26 which corresponds to approximately two standard deviations ( α = 2 . 3% ) above the sample mean of the negative control . We set the minimum read threshold in the +TAP asynchronous library to be 25 . Using RNA-seq data from [13] we selected only TSSs that had more than a 35% increase in the downstream RNA-seq coverage ( S1 Dataset ) . In total , this procedure identified 2 , 726 Caulobacter TSSs ( S1 Dataset ) . The parameters we set for TSS identification allowed us to identify the TSSs for approximately 75% of genes and operons , in keeping with similar estimates of the percentage of TSSs reported for Listeria monocytogenes [17] and Escherichia coli [18] . Two factors are likely to contribute to the null identification of TSSs: strict parameter cutoff based on TAP-enrichment ( S1 Fig . ) and 5′ RACE dependence on a 5′ ligation , which is inefficient for RNAs that are highly structured at the 5′ region [19] . To verify our approach , we tested 36 of the TSSs identified by 5′ global RACE using β-galactosidase reporter assays and found that all 36 exhibited significant expression activity ( S3 Dataset ) . Canonical bacterial promoters contain binding sites for σ factors upstream of the TSS . A motif search of 50 bp upstream of the identified TSSs revealed a−35 ( TTG ) and −10 ( A/T ) binding site ( n = 1 , 666 ) consistent with σ73 ( RpoD ) , the most abundant housekeeping sigma factor in Caulobacter [20] ( S2 Fig . , S4 Dataset ) . The 5′ nucleotide of 93% ( 1 , 542/1 , 666 ) of the identified RpoD binding motifs are positioned between −34 to −37 bp upstream of the TSSs . Based on recent functional re-annotations of the Caulobacter genome ( CP001340 ) using ribosome profiling and computational analysis [13] , [21] , we categorize TSSs into four categories with overlap ( Fig . 1B ) . TSSs located upstream of CDS are denoted as primary ( P , 1443 ) ; those located in intergenic regions or upstream of annotated RNAs are denoted as non-coding ( N , 155 ) ; those initiated from within coding sequences and transcribed in the same direction are denoted as internal ( I , 344 ) , and those transcribed in the opposite direction of the CDS are denoted as antisense ( A , 503 ) . There is overlap between primary and antisense TSSs ( A+P , 84 ) and between primary and internal TSSs ( I+P , 197 ) . Thirty-three of the 93 previously characterized non-disruptable intergenic gaps in the Caulobacter genome [22] were found to contain a TSS within the non-disruptable gap , suggesting these TSSs may play a role in cell viability . We also observed a slight directional bias in the number of TSSs encoded in the same direction as DNA replisome movement ( n = 1544 co-directional , n = 1182 opposing ) ( Fig . 1C ) to minimize collisions between DNA polymerase and RNA polymerase during chromosome replication [23] . Comparison of global TSS levels of the swarmer and stalked cell stage of the cell cycle revealed differences in the pattern of global site-specific TSS levels ( Fig . 1D ) . We analyzed the direction of cell cycle-regulated TSSs peaking in the swarmer cell and find no significant directional bias ( n = 21 co-directional , n = 19 opposing ) . As the swarmer cell does not actively replicate its chromosome , it is likely that these promoters have no selective pressure to be encoded co-directional with DNA replication . To distinguish cell cycle-regulated TSSs from constitutively active TSSs , we implemented a modified Fourier Transform algorithm , similar to [24] , including both minimum sequence read and expression fold-change cutoffs on the corresponding normalized time-course sequencing data of the 2 , 726 identified TSSs . Using this approach , we identified cell cycle-regulated TSS levels for 586 TSSs ( Fig . 2A ) . In general , the cell cycle TSS levels measured by 5′ global RACE yield similar timing as the steady state mRNA levels as determined previously by microarrays [25] ( S3 Fig . ) . To improve upon lower resolution and coverage studies of transcription factor binding motifs [25] , we used our base pair resolution TSSs and a new genome annotation [13] , [21] to search in DNA segments upstream of cell cycle-regulated TSSs for binding motifs of the core cell cycle circuit ( Fig . 2B ) : CtrA ( Fig 2C , S4 Fig . , S5–S7 Dataset ) , SciP ( Fig . 2C , S4 Fig . , S8–S9 Dataset ) , DnaA ( Fig . 2C , S4 Fig . , S10 Dataset ) and the CcrM DNA methyltransferase ( Fig . 2C , S4 Fig . , S11 Dataset ) . To identify TSSs regulated by the GcrA transcription factor , we searched upstream of TSSs for DNA segments enriched in GcrA ChIP-seq signal ( S12 Dataset ) [26] . About 57% of cell cycle-regulated TSSs had upstream binding motifs for one or more of the four transcription factors or CcrM DNA methlytransferse in the core cell cycle regulatory circuit ( Fig . 2C , S4 Fig . , S5–S12 Dataset ) . We identified 199 cell cycle regulated TSSs with a single upstream regulatory binding motif for one of the known master regulators ( DnaA , 29; GcrA , 34; CtrA , 89; SciP , 7; CcrM , 40 ) . In addition , we found another 135 TSSs that have multiple master regulatory factor binding sites , suggesting they are under combinatorial control ( S13 Dataset ) . The TSSs that are preceded by multiple regulatory motifs are enriched for genes encoding critical cell cycle proteins , including the genes of the core regulatory circuit itself . We also identified binding motifs for sigma factors SigT ( S5 Fig . , S14 Dataset ) and RpoN ( S5 Fig . , S15 Dataset ) . The TSSs with binding motifs for SigT , a cell cycle-regulated ECF sigma factor [11] that is also induced under stress [27] , exhibited peak levels during the swarmer to stalked cell transition coincident with the expression pattern of sigT ( S5 Fig . ) . The previously identified Caulobacter SigT binding motif ( GGAAC-N16-CGTT , e-value = 1 . 9 e−39 , n = 26 ) [25] , [27] is located at the −35 bp region relative to the TSS ( S5 Fig . ) . RpoN , a sigma factor induced upon nitrogen limitation and required for flagella gene expression in Caulobacter [28] , [29] , controls two classes of genes in both the swarmer to stalk transition and in flagellar genes . The previously identified RpoN binding motif ( GGCNC-N4-CTTGC , e-value = 1 . 5 e−27 , n = 33 ) [25] , [30] is located between −35 bp and −25 bp relative to the TSS ( S5 Fig . ) . The RpoD , SigT , and RpoN binding motifs together account for 63% ( 1725/2726 ) of the observed upstream TSSs regions . The remaining TSSs are likely activated by the additional 13 known sigma factors encoded in the Caulobacter genome . The CtrA response regulator is a master transcriptional regulator of the Caulobacter cell cycle that was shown to directly control 95 cell cycle-regulated genes using ChIP-chip [11] , [12] . We identified 183 cell cycle-regulated TSSs with an upstream CtrA binding motif ( Fig . 3A ) that were also enriched in CtrA ChIP-seq data [31] ( S5–S7 Dataset ) . Among these are promoter regions of the cell cycle master regulators sciP , ccrM , and ctrA , a regulator of CtrA degradation rcdA , cell division proteins ftsK , ftsQ , and mipZ , the response regulator divK , flagellar genes , and 6 non-coding RNAs ( S5–S7 Dataset ) . Surprisingly , we observed two classes of CtrA binding motifs , a full palindromic TTAA-N7-TTAA ( Fig . 3A ) and a half motif TTAA ( Fig . 3A ) . Based on hierarchical clustering of the cell-cycle profiles we identified 3 groups of CtrA binding motifs ( CtrA full , CtrA half repressor , and CtrA half activator ) . Expression of genes controlled by CtrA full ( n = 52 , S5 Dataset ) mirrored CtrA protein levels in the predivisional cell stage ( 60–120 min ) ( Fig . 3A ) . The 5′ nucleotide of these motifs is positioned near the −35 region , consistent with CtrA activity as a transcriptional activator in the predivisional cell . Conversely , CtrA half repressor containing promoters ( n = 24 , S6 Dataset ) exhibited an anti-correlated temporal pattern of TSS activation with the CtrA protein levels ( Fig . 3A ) . These half sites were positioned over the −10 site , consistent with CtrA functioning as a transcription repressor , similar to the observed repression of ctrA P1 by CtrA [32] ( Fig . 2B ) . Fifty eight percent ( 14/24 ) of promoters with a CtrA half repressor motif also contain a DnaA binding motif or a GcrA binding site . CtrA half site-containing promoters that function as activators ( n = 107 , S7 Dataset ) correlated with CtrA protein levels throughout the cell cycle , with transcription activity occurring in both the swarmer and predivisional cell stages . These CtrA half activator binding sites are also near the −35 region , consistent with transcriptional activation ( Fig . 3A ) . Indeed , gene expression profiling studies using microarrays from strains with altered CtrA activity [28] show good agreement with CtrA full and CtrA half activator in transcription activation and CtrA half repressor in transcriptional repression ( S5–S7 Dataset ) . These data show that the activity and timing of transcription is controlled by the precise position of CtrA DNA binding upstream of the TSS . Additionally , two transcription factors ( MucR 1/2 ) that act to regulate the S-G1 phase transition were reported to bind to CtrA target promoters . We found that 76% of cell cycle-regulated promoters under MucR 1/2 ChIP-seq peaks , as determined by the Fumeaux et al . [33] , contains a CtrA binding motif , and 10% of the cell cycle-regulated promoters under MucR 1/2 ChIP-seq peaks contain a SciP binding motif . Those CtrA regulated promoters under MucR 1/2 peaks were more highly repressed in the stalk/early-predivisional cells than CtrA-regulated promoters without MucR 1/2 ( S10 Fig . ) . This is consistent with the proposal by Fumeaux et al . [33] that MucR 1/2 specifically represses CtrA activated genes in the S phase while SciP specifically represses CtrA activated genes in the swarmer cell . There are a total of 61 TSSs with a SciP binding motif [8] in the upstream promoter region , and among these are the promoter regions of ctrA , the DNA methyltransferase ccrM , and polar development protein podJ . Previously , the promoter regions of 30 genes were identified as targets of SciP by expression arrays [7] , 15 of which were shown to be direct by ChIP-PCR [8] . Seven of the 15 promoter regions shown to directly bind SciP , were identified by the TSS motif search . As with the CtrA binding motifs , we found that the SciP motif falls into two categories: a palindromic motif GCGNC-N5-GNCGC and a half motif GCGAC ( Fig . 3B ) that was identified previously ( reverse complement in [8] ) . TSSs with the palindromic motifs exhibited peak levels at 120 min ( n = 29 , S8 Dataset ) and those with half SciP motifs exhibited levels peaking at 100 min ( n = 32 , S9 Dataset ) ( Fig . 3B ) . Both groups exhibited an anti-correlated cell cycle profile with SciP protein levels indicating SciP acts as a repressor , in agreement with previous reports [7] , [8] . To confirm this role as a repressor , we showed that the mutation of the SciP site in three CtrA activated promoters leads to an increase in the promoter activity ( S9 Fig . ) . The half motif is associated with early TSS repression and the full motif with repression later in the cell cycle ( Fig . 3B ) . SciP binding motifs are predominantly found between −60 bp and −90 bp relative to the TSS ( Fig . 3B–C ) and 80% ( 49/61 ) of TSSs with a SciP binding motif also have a CtrA binding motif ( Fig . 3C ) . On average the SciP ChIP-seq signal peak occurs upstream of the CtrA ChIP-seq signal in agreement with the upstream position of SciP binding motifs relative to CtrA ( S8 Fig . ) . When SciP sites are combined with CtrA sites , SciP represses genes in the late predivisional cell and the swarmer cell ( Fig . 3C ) where SciP protein levels peak ( Fig . 3B ) . The DnaA protein directs the initiation of chromosome replication in addition to functioning as a transcription factor [10] . We identified DnaA binding motifs in 77 promoter regions of cell cycle regulated TSSs ( S10 Dataset ) . DnaA regulates transcription of GcrA , FtsZ , PodJ , and components of the replisome and nucleotide biosynthesis proteins [10] . The DnaA binding motif occurs as the sole master regulator site in 29 promoter regions , while it is commonly accompanied at promoter regions by 25 CtrA , 19 CcrM , and 19 GcrA sites ( Fig . 2C , S4 Fig . ) . The GcrA protein is a master transcription factor that is activated by DnaA ( Fig . 2B ) [34] and whose protein levels are anti-correlated with CtrA [9] . We searched for enrichment of the GcrA signal in promoter regions of cell cycle regulated TSSs in the ChIP-seq dataset reported by [26] and found GcrA binding sites in promoters of CtrA P1 , podJ , and mipZ in addition to 91 other promoter region ( S12 Dataset ) . GcrA binding occurs as the sole master regulator site in 34 promoter regions , while it is accompanied by 30 CtrA , 29 CcrM , and 19 DnaA binding motifs ( Fig . 2C , S4 Fig . ) . Hemi-methylated GANTC sites in the Caulobacter chromosome are recognized by the CcrM DNA methyltransferase yielding 6-methyl adenines [35] , [36] . The transcription of CtrA and DnaA is affected by the cell cycle-dependent methylation state of their promoters , a link that helps synchronize the progression of the core cell cycle regulatory circuit with the progress of DNA replication [3] , [37] , [38] . We identified a total of 96 TSSs with GANTC sites located within 50 bp upstream of the TSS that were activated at specific stages in the cell cycle ( Fig . 4A–B , S11 Dataset ) . Eleven of the 96 TSSs contain more than one upstream GANTC site yielding a total of 108 GANTC sites within 50 bp upstream of cell cycle regulated TSSs . These cell cycle-regulated TSSs fell into three distinct temporal clusters . For the TSSs with the GANTC site positioned between −10 and −35 of the promoter region , the time of TSS activation occured between 40 and 60 minutes ( Fig . 4B–C red ) . If the GANTC motif is positioned outside this region the TSS levels is lowest between 40 and 60 minutes ( Fig . 4B–C blue ) . A third cluster of cell cycle-regulated TSSs that have GANTC sites equally distributed within 50 bp upstream of the TSS exhibits maximal levels between 80 to 100 minutes ( Fig . 4C green ) . Fifty-eight percent of all TSSs with upstream GANTC motifs contained other master regulator binding motifs ( Fig . 2C , S4 Fig . ) . Based on this TSS study and a recent RNA-seq study [13] , we identified 587 ( 503 A , 84 A+P ) antisense transcripts in the Caulobacter genome . Only eight antisense transcripts for genes encoding transposases have been reported previously [39] . We found an additional 179 putative antisense TSSs that have RNA-seq coverage [13] below our mapping threshold ( S16 Dataset ) . Despite the low RNA-seq coverage , 7 ( out of 7 assayed ) antisense TSS with low RNA-seq coverage had significant β-galactosidase activity when 75 bp of the promoter were inserted in front of the β-galactosidase gene ( S3 Dataset ) , suggesting they are indeed antisense TSSs . 583/3 , 885 ( ∼16% ) of Caulobacter CDSs have at least one antisense TSS; as compared to Helicobacter pylori ( 27% ) [40] , Escherichia coli ( 20% ) [41] , [42] , and Mycoplasma pneumoniae ( 12% ) [43] . Seventy-four antisense TSSs reside within essential genes including those that encode DnaA , CtrA , an RNA polymerase beta chain , and MreB . Of the 766 ( 587+179 ) antisense TSSs in the Caulobacter genome , 107 are cell cycle regulated ( Fig . 5A ) . Of these 107 , 42 are within genes that are constitutively expressed and 13 are within genes that are cell cycle controlled . We observed for the spmX gene , that promotes the localization and activation of the DivJ histidine kinase [44] , the timing of antisense transcription is correlated with the timing of sense transcription ( Fig . 5B ) . Perhaps the antisense transcript stabilizes the spmX mRNA , as reported for the gadX mRNA by the antisense GadY transcript in E . coli [45] . In 12 of the 13 cell cycle-regulated antisense TSSs residing within with a cell cycle controlled gene ( S17 Dataset ) , the levels of the antisense TSS and the corresponding cell cycle-regulated primary TSS peak at different times over the course of the cell cycle , as shown for CCNA_01391 ( Fig . 5C ) . The antisense TSS with a SigT binding site in its promoter is induced at the swarmer to stalked cell transition . Upon the decrease in levels of the antisense TSS , we observe an increase in the levels of the primary TSS ( Fig . 5C ) . The coordinated transcriptional patterns of these genes and their antisense TSSs , suggest that the antisense RNA may control gene expression . Bacterial intergenic small non-coding RNAs ( ncRNAs ) have been shown to enable cells to adapt to environmental and physiological challenges [46] . We have separately reported 199 ncRNAs in the Caulobacter genome [13] , including four new ncRNAs that are encoded in nondisruptable regions of the genome [22] . We identified 155 TSSs within intergenic non-coding regions ( category N , Fig . 1B ) using 5′ global RACE in minimal medium; these included 50 TSSs for tRNA or rRNA genes . In 46 of these 155 intergenic TSSs , the TSS matches the 5′ nucleotide in the ncRNA identified by RNA-seq [13] ( S18 Dataset ) . While only 5 Caulobacter ncRNAs were previously observed to be cell cycle-regulated [39] , [47] , we identified 33 cell cycle-regulated non-coding TSSs activated in different phases of the cell cycle ( Fig . 6A , S1 Dataset ) . One of these cell cycle-regulated TSS drives a ncRNA of 182 nt in length ( CCNA_R0094 ) transcribed from within the chromosomal origin of replication ( Fig . 6B ) that appears to be essential [22] . We identified a total of 241 CDSs with multiple upstream TSSs ( S19 Dataset ) that appear to be independently controlled . Fifty seven of these CDSs containing multiple promoters are essential for viability [22] . In Caulobacter , only 18 cell cycle-regulated genes , including ctrA , dnaN , clpX , and rcdA , have been shown previously to be transcribed from multiple cell cycle controlled promoters using either tiling microarrays , nuclease protection , or primer extension assays [25] , [32] , [48] , [49] . We found that 102 CDSs ( 42% of those with multiple upstream TSSs ) have at least one cell cycle-regulated TSS , and 25 CDSs , including ctrA ( 3 promoters , Fig . 7A ) and podJ ( 2 promoters , Fig . 7B ) , have more than one cell cycle-regulated primary TSS that are independently regulated . The ctrA gene was previously shown to be transcribed from two promoters ( P1 and P2 ) where P1 is activated by GcrA after the ctrA locus is replicated and P1 becomes hemi-methylated ( Fig . 2B ) [9] , [37] . The P1 promoter is thereafter repressed by CtrA which simultaneously strongly activates P2 [32] . We confirmed the previously reported temporal sequence of P1 and P2 activation , and identified an additional cell cycle-regulated promoter , P3 , located between P1 and P2 ( Fig . 7A ) . A CtrA half repressor binding motif TTAA is located −14 bp upstream of the P1 TSS , in addition to a SciP binding motif GCGAC located −78 bp upstream , and a CcrM methylation site GANTC ( ▾ ) located at −28 bp upstream of P1 ( S6 , S9 , S11 Dataset ) . A full CtrA binding motif TTAA-N7-TTAA is located at −39 bp upstream of P2 and at −14 bp upstream of P3 , and a half SciP binding site is located at −74 bp upstream of P3 ( S5 , S9 Dataset ) . Both SciP and CtrA have been shown previously to bind at both locations [8] , [32] . The full CtrA motif likely functions to activate P2 and simultaneously repress P3 because the 5′ nucleotide of the CtrA full motif is at −14 bp upstream of P3 ( repression ) and at −39 bp upstream of P2 ( activation ) . PodJ is an essential protein that mediates polar organelle development by contributing to the synthesis of pili and the control of the polar localization of the PleC kinase/phosphatase [50] , [51] . Two cell cycle-regulated primary TSSs ( P1 and P2 ) are located 120 bp and 22 bp upstream , respectively , of podJ ( Fig . 7B ) . The levels of P2 , which contains a CcrM methylation site GANTC ( ▾ ) −24 bp upstream ( S11 Dataset ) and a DnaA binding motif ( CTCCACA ) ( Hottes et al , 2005 ) at −82 bp upstream ( S10 Dataset ) , peaks at 40 min into the cell cycle during the swarmer to stalked cell transition . P1 contains a CtrA full binding motif TTAA-N6-TTAA at −49 bp upstream ( S5 Dataset ) and a SciP half binding motif GCGAC at −73 bp upstream ( S9 Dataset ) with P1 levels peaking at 100 min into the cell cycle in the predivisional cell , contributing to a second wave of podJ transcription . For 77 out of the 241 CDSs transcribed with multiple primary TSSs one promoter is constitutively expressed while the other is cell-cycle regulated . For example mipZ ( Fig . 7C ) , which encodes the essential division plane positioning ATPase MipZ [52] , has a constitutive P1 and a cell cycle-regulated P2 promoter containing a CtrA half repressor motif at −10 bp upstream of the TSS ( Fig . 7C , S6 Dataset ) . Activation of the P2 promoter results in an increase in the transcription of MipZ at the same time as transcription of FtsZ , thereby coordinating the temporal control of MipZ regulation of FtsZ function [6] . In many instances , the transcription of individual genes within the 848 operons is differentially regulated . There are 115 operons that have a cell cycle-regulated TSS upstream of the leading CDS . There are 52 operons ( S20 Dataset ) with internal cell cycle-regulated TSSs for downstream CDSs enabling independent cell cycle regulation of downstream operon genes . One example is the operon consisting of CCNA_00875 , CCNA_00876 , and CCNA_00877 , where CCNA_00875 encodes a 7 , 8-dihydro-8-oxoguanine-triphosphatase , CCNA_00876 encodes a Flp/Fap pilin component protein , and CCNA_00877 encodes a protein of unknown function ( Fig . 7D ) . In this operon , the P1 start site is constitutively active , but the P2 start site has a putative CtrA full binding motif TTAC-N7-TTCA upstream of it ( S5 Dataset ) where the 5′ nucleotide of the motif resides at -39 bp and a CcrM methylation site GANTC ( ▾ ) is at −35 bp ( S11 Dataset ) . The P2 TSS has a cell cycle-dependent profile suggesting cell cycle-regulated expression of just the downstream CDSs in the operon .
Genome-wide assays have shown that antisense transcription occurs in many bacteria species [40] , [41] , [43] . Further , antisense transcripts have been shown to regulate genes involved in a wide variety of processes such as photosynthesis in Synechocystis PCC6803 [54] , acid and SOS response in E . coli [45] , [55] , virulence in Salmonella enterica [56] , and iron transport in Vibrio anguillarum [57] . The 107 cell cycle-regulated antisense TSSs suggest that antisense regulation is a significant element of cell cycle regulation . Supporting this , we found that 23 of the 107 cell cycle regulated antisense TSSs have binding sites in their promoter regions for the core cell cycle regulatory factors . Antisense TSSs are also found within the CDSs of essential cell cycle-regulated genes , including dnaA , ctrA , spmX and mreB . Most previously described mechanisms of antisense transcripts involve base-pairing with the corresponding sense mRNA to alter the RNA stability [45] , [54] , translation [58] , or transcription termination [57] . Depending on the context , the antisense RNA has been observed to effect either down-regulation or up-regulation of genes . In Caulobacter , for 12 cell cycle regulated primary TSSs , the TSS levels is anti-correlated with the time of the time of activation of the antisense TSS ( S17 Dataset ) , suggesting that antisense transcripts may down-regulate the levels of their target mRNAs . Additionally , the Caulobacter antisense TSSs show multiple cell cycle expression patterns , suggesting that this mechanism is active in regulation of gene expression during all stages of the cell cycle . Twenty four promoters upstream of the 107 cell cycle-regulated antisense TSSs contain master cell cycle regulator binding motifs allowing them to be controlled directly by these cell cycle regulating transcription factors . Antisense TSSs are also abundant in Sinorhizobium meliloti [59] , another α-proteobacteria . However , as master regulator binding motifs residing within the protein coding sequences have generally not been included in global analyses of the α-proteobacterial cell cycle transcription control circuitry [25] , [60] , [61] and it is not known whether antisense TSSs controlled by the cell cycle master regulators are conserved within the α-proteobacteria . In Caulobacter , 199 intergenic ncRNAs have been identified in the genome including rRNAs and tRNAs [13] , [39] . Of these , we identified 155 TSSs for ncRNAs ( S1 Dataset ) and for 33 of these the TSS levels are cell cycle-regulated ( Fig . 6A ) . The functions of only two ncRNAs ( non rRNA/tRNA/RNaseP/4 . 5S RNA ) have been characterized: the tmRNA which rescues stalled ribosomes and alters the timing of replication initiation , and crfA which controls the carbon starvation response [47] , [62] , [63] . While ncRNAs perform many functions , a common ncRNA function in bacteria involves ncRNA base-pairing to mRNAs to regulate gene expression , sometimes mediated by the RNA chaperone Hfq [46] , [64] . Four of the ncRNAs lie within non-disruptable intergenic gaps [22] suggesting that these ncRNAs may be essential for the regulation of Caulobacter cell cycle progression . Across the Caulobacter genome we identified 241 CDSs with multiple upstream TSSs ( S19 Dataset ) including 57 of essential [22] and 102 of cell cycle regulated genes . The genes encoding the CtrA global cell cycle regulator and the PodJ polar differentiation factor have multiple upstream promoters with different cell cycle timing that modulates the pattern of expression of the genes . In the case of ctrA transcription , GcrA activates the hemi-methylated P1 promoter to initiate ctrA transcription followed by a boost in CtrA production from the subsequent auto-regulated activation of the P2 promoter [9] , [32] ( Fig . 2B , Fig . 7A ) . Based on the identification of a third temporally controlled promoter , P3 , whose temporal pattern of activation is similar to that of P1 , we suggest that P1 and P3 accelerate initial production of CtrA . Both P1 and P3 have a CtrA binding site in the −10 region upstream of each TSS , which are then repressed by CtrA . We also know , as mentioned earlier , that the subsequent expression of SciP represses ctrA transcription . The net effect appears to be aimed at modulating the shape of the pulse of CtrA production to make it stronger , yet shorter in time . In the case of podJ , the function of the additional promoters seems to be to extend of the duration of PodJ production over a longer interval of the cell cycle . In other cases , such as the cell division gene mipZ , a cell cycle-regulated promoter is activated at a specific time during the cell cycle , corresponding to the time in which the protein product is needed , but the gene is also transcribed at a low level from a constitutive promoter , presumably ensuring that a low level of MipZ is always present during the cell cycle ( Fig . 7C ) . We found that 209 operons contain internal TSSs , 55 of which are cell cycle-regulated ( S20 Dataset ) . In some operons , downstream genes are activated at different times in the cell cycle . Internal TSSs that exhibit independent cell cycle-regulated expression have nearby upstream transcription factor binding motifs . Additionally , we report separately that some promoters internal to operons lead to production of alternative shortened forms of the encoded protein [13] , presumably changing the protein's function . The spatial ordering of multiple upstream promoters , in conjunction with promoters internal to operons , could conceivably have a regulatory impact since each mRNA has a different 5′ UTR sequence that would enable differential post-transcriptional control . The exciting implications of the combinatorial promoter logic possible for genes and operons with multiple TSSs remain to be explored . The largest class of cell cycle-regulated TSSs are those that have one or more CtrA binding motifs in the promoter region ( 183 of the 586 of the cell cycle-regulated TSSs ) ( Fig . 2C , 3A , S4 Fig . ) . We observe two CtrA binding motifs , a full palindromic binding site and a half palindromic binding site ( Fig . 3A ) . We find that the 5′position of the CtrA full palindromic motif relative to the TSS is most commonly at the −35 region corresponding to activation of transcription in the predivisional cell . Conversely , the 5′position of the CtrA half-palindromic motifs can either function as a repressor by binding over the −10 , or an activator by binding over the −35 region . In predivisional cells , CtrA full-palindromic TSSs maximally activate at 80 to 100 min while those with CtrA half activator , do so later at the 120 min time point , likely due to the tighter binding affinity of the CtrA full site to CtrA∼P [65] . TSSs with CtrA half motifs are also active in the swarmer cell , while CtrA full motif TSSs are only active in the predivisional cell . We interpret this switch in TSS levels to be primarily due to overlapping control by SciP which inhibits CtrA activated promoters as reported by [7] , [8] . About 80% ( 49/61 ) of TSSs containing an upstream SciP motif also have a CtrA motif ( Fig . 2C , 3C , S4 Fig . ) . The SciP binding motifs are positioned between −60 and −90 upstream of the TSSs ( Fig . 3B , S8 Fig . ) . SciP binds directly to DNA as shown in [8] and confirmed here by mutation of SciP sites . We have mutated both the half and full SciP sites in 3 CtrA activated promoters and in each case we observed an increase in promoter activity as measured by β-galactosidase activity , providing evidence that the SciP motif does indeed function to bind SciP and acts that bound SciP is a repressor of these promoters ( S9 Fig . ) . In the presence of DNA containing both CtrA and SciP binding motifs , both CtrA and SciP become resistant to proteolysis [66] . Since direct interaction between CtrA and SciP has been demonstrated [7] , [8] , it seems likely that the combined DNA binding energy provides additional regulatory capacity . Interestingly , SciP and CtrA binding motifs are primarily positioned together with at least one motif present as a full palindromic binding site ( 91 . 8% of co-regulated promoters ) and the co-positioning of SciP and CtrA half motifs occurs only in four co-regulated promoters . It is possible that SciP and CtrA require sufficient binding to DNA to form a stable complex that is not accomplished with weaker half sites . A recent report by Fumeaux et al . [33] analyzed the top 50 SciP ChIP-seq peaks and found them to contain TTAACAT motifs , similar to the CtrA half binding motif . We performed motif searches of the SciP ChIP-seq peaks reported by Fumeaux et al . [33] using the CtrA half- and SciP half-binding motif presented in this paper , and found a total of 63 CtrA half motifs and 143 SciP half motifs with a P value less than 1−3 . Both Fumeaux et al . [33] and our own analysis of their data revealed that only a subpopulation of SciP ChIP-seq peaks ( 47% ) contain the TTAACAT motif . However , we found that 88% of the SciP ChIP-seq peaks contain a SciP binding motif . Of the 47% that contain a CtrA binding motif , 33/35 of these also contain the SciP binding motif . The SciP motif used in this study is based on previously reported direct footprints to DNA , ChIP-chip analysis and microarray analysis reported in Tan et al . [8] , which is in agreement with a motif previously identified for a cohort of genes expressed at the same time in the cell cycle by McGrath et al . [25] . The SciP binding motif correlates with a larger percentage of the SciP ChIP-seq peaks than the CtrA motif . Both SciP half and CtrA half ( TTAACAT ) motifs are present in the CtrA promoter foot-printed by Tan et al . [8] . However , only protection of the SciP sites was observed with purified SciP . We suggest that the Fumeaux et al . ChIP-seq data analysis [33] missed many of the SciP motifs in the presence of stronger CtrA motif signals from peaks including SciP and CtrA co-regulated promoters . Indeed , the peaks of the Fumeaux et al . [33] SciP and CtrA ChIP-seq signal match the positions of their respective binding motifs ( S8 Fig . ) . We found that 57% of the cell cycle-regulated TSSs have upstream binding sites for known cell cycle transcriptional regulators ( Fig . 2B-C , S4 Fig . , S5-S12 Dataset ) whose activity and protein levels oscillate in time [1] , [4] . While 199 cell cycle regulated TSSs contain a single regulatory factor binding site , 135 have binding sites for 2 or more of these factors ( S13 Dataset ) . We have not yet identified the regulatory factors ( and their binding sites ) that control the other 43% of the cell cycle regulated TSSs but we expect to find that the activity of these promoters are controlled by a second layer of regulatory factors that are turned on by the core circuit ( Fig . 2B ) . Co-regulation among the 5 cell cycle master regulators acts to tune the timing of cell cycle transcription profiles , and we expect to find equally complex timing regulation in every genetic regulatory sub-system . DnaA directly controls the initiation of DNA replication by binding to the origin of replication [67] and it also serves as a transcription factor for a large complement of cell cycle regulated genes [10] . We identified 77 cell cycle-regulated TSSs with a DnaA binding motif . Additionally , from our analysis of the ChIP-seq data of [26] we found 94 cell cycle-regulated TSSs with enriched upstream binding of GcrA , a transcriptional activator [9] . In contrast to to the temporal expression of CtrA-controlled genes , DnaA and GcrA regulated TSSs have a multitude of different cell cycle profiles with lower levels of activation ( S6 Fig . ) . This increase in profile diversity is likely due to an increased number of co-regulated promoters for genes controlled by both DnaA and GcrA . Additionally , when DnaA and GcrA binding sites are combined with a CtrA binding site , the transcriptional profile appears to be dominated by the activity of CtrA . Finally , we identified CcrM methylation sites within 96 cell cycle-regulated TSSs ( Fig . 4A–B ) . We found cell cycle-controlled TSSs with GANTC motifs in these promoter regions , such as the ctrA P1 promoter which is activated by hemi-methylation , in agreement with a previous report of the control of this gene by methylation state of its promoters [37] . A cluster of these TSSs , located between −35 and −10 of the promoter region ( Fig . 4B–C Red ) , are activated between 40 and 60 minutes . Another cluster of TSSs , with GANTC sites positioned outside of the −10 and −35 region , exhibit minimal TSS levels between 40 and 60 minutes ( Fig . 4B–C Blue ) . While the underlying mechanism of methylation dependent transcription regulation in Caulobacter is unknown , the GcrA transcription factor has been implicated in the control of some GANTC containing promoters [68] . However , we find that only 31% of cell cycle regulated TSSs enriched for GcrA binding ( from the published ChIP-seq data from [26] ) also contain GANTC methylation sites within 50 bp upstream of the TSS ( Fig . 2C , S4 Fig . ) . Additionally , the promoters of only 11 genes found to bind GcrA by ChIP-seq are differentially expressed in GcrA or CcrM depletion arrays [69] . Cumulatively , the newly-revealed complexity of transcriptional regulation that drives the Caulobacter cell cycle , coupled to multiple modes of post transcriptional regulation [13] has opened up new avenues of bacterial systems control architecture . The known Caulobacter cell cycle regulatory circuit , composed of four transcription factors and a DNA methyltransferase , drives stage specific expression of a majority of cell cycle-regulated promoters [2] . We found multiple examples of co-regulation using these five master regulators of the cell cycle regulatory circuit , and many of these examples involve essential genes and essential regulatory regions of the genome [22] ( S13 Dataset ) . The presence of each master regulator is restricted to specific times in the cell cycle [8] , [70] , so that there are multiple cell cycle stage-specific regulatory options to tune the timing of transcription . The core Caulobacter cell cycle transcription regulatory network is conserved in a majority of α-proteobacteria [2] , [60] and has been shown to be integrated with other regulatory pathways such as quorum sensing [71] , plant symbiosis [61] , [72] , gene transfer agent production [73] , [74] , host cell infection [75] , [76] , and motility [73] . These α-proteobacterial species have adapted their cell cycle control circuits to their specific biological niches . Negative feedback by antisense regulation and the existence of many promoters with multiple TSSs that appear to modulate the cell cycle stage of mRNA production identifies yet other levels of regulatory control .
All 5′ Global RACE experiments were performed with C . crescentus strain CB15N ( NA1000 ) [77] . Cells were grown to OD600 0 . 4 in M2G [78] and synchronized using standard procedures [77] . Aliquots were taken at 20 min intervals over ∼1 cell cycle ( 140 min ) , pelleted and immediately frozen in liquid nitrogen . Total RNA was isolated using Trizol Plus kit ( Invitrogen ) according to the manufacturer protocol . Ten micrograms of total RNA isolated from each aliquot of cells were treated with MICROBExpress ( Ambion ) to remove ribosomal RNA following the manufacturer protocol . Ribosomal RNA depleted total RNA was treated with 10 U of Tobacco Acid Pyrophosphatase ( TAP ) ( epicentre ) for 1 . 5 hours at 37°C . In addition , two sequencing libraries were prepared from an unsynchronized cell population grown to OD600 0 . 4 in M2G , one of which was not treated with TAP . Purified RNA was ligated with 10 pmol of RNA adapter ( 5'-ACACUCUUUCCCUACACGACGCUCUUCCGAUCU-3' ) using 25 U of T4 RNA Ligase 1 ( New England BioLabs ) for 12 h at 16°C . cDNA was then generated from purified RNA using SuperScript II Reverse Transcriptase ( Invitrogen ) with 10 pmol of primer ( 5'-CTCGGCATTCCTGCTGAACCGCTCTTCCGATCTNNNNNN-3' ) following standard manufacturer protocol then treated with 1 Unit of RNase H ( Invitrogen ) for 20 min at 37°C . cDNA was PCR amplified for 12 cycles using primers ( 5'-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT-3' , 5'-CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT-3' ) . DNA was size selected from 100 bp to 300 bp using agarose gel-electrophoresis and purified using QIAquick Gel Extraction Kit ( Qiagen ) . 100ng of DNA was then treated with duplex-specific nuclease ( evrogen ) to remove ribosomal cDNA following standard manufacturer protocol and PCR amplified for 10 cycles . Sequencing reads were mapped using Bowtie version 0 . 12 . 7 [16] to the C . crescentus NA1000 reference sequence ( Genbank: NC_011916 ) . All valid alignments were made , and only reads that mapped to a unique location were used with no mismatches allowed . Sequencing reads were normalized by the total number of non-ribosomal RNA reads for comparison . +TAP and -TAP reads were normalized with relation to each other , and reads from synchronized cells were normalized with relation to each other . To differentiate between reads initiating from TSS and RNA processing sites , thirty-four previously biochemically characterized TSS with a minimum normalized read value of at least 30 in the +TAP library were used as a positive control , and twenty-four tRNA processing sites ( n = 24 ) with a minimum normalized read value of at least 30 in the -TAP library were used as a negative control . The natural log value of the ratio ( θ ) between the number of sequencing reads obtained at the 5′ ends of positive and negative controls in +TAP/-TAP libraries was calculated . The mean and standard deviation of θ obtained for positive and negative controls were obtained ( Welch two sample t-test , df = 41 . 2 , p-value = 1 . 3 e−11 , S1 Fig . ) . To minimize false-positives or Type I errors in TSS prediction , we set α = 2 . 3% , corresponding to θ>0 . 26 . An increase of >35% of RNA-seq coverage [13] in a 38 bp window upstream compared to downstream of the TSS was also required . The distance of mapped 5′ read locations upstream to CDSs in the Caulobacter genome was then determined . Reads that are mapped within 300 bp upstream of annotated CDSs or on the first base of the start codon were categorized as upstream promoters ( category “P” ) ; those within 300 bp upstream that are inside the upstream CDS but reside within the last 30% of the CDS if the CDS is longer than 600 bp or within the last 30% of the CDS if the CDS is shorter than 600 bp are categorized as internal upstream promoters ( category “IP” ) . If the upstream CDS is oriented in the opposite direction as the TSS , then the TSS is categorized as an antisense upstream promoters ( category “AP” ) . Unless already categorized as IP or AP for another CDS , reads that are mapped inside CDSs are categorized as internal promoters ( category “I” ) ; if oriented in the opposite direction of the CDS , the TSS is categorized as an antisense promoter ( category “A” ) . All other reads were categorized as non-coding TSS ( category “N” ) . The total number of reads corresponding to each CDS as obtained from the +TAP library was calculated for both the plus and minus strands . Upstream promoter locations categorized as P , AP , or IP with reads contributing more than 10% of the total reads in the same orientation for a single CDS were kept . Antisense or internal promoter locations ( A or I ) with reads contributing more than 25% were kept . If multiple reads were mapped within a distance of ≤5 bp , the reads were summed and placed at the position with the highest number of reads in the +TAP library . TSS with ≥2 normalized reads in at least one of the cell cycle time points were kept for categories ( P , AP , or IP ) ; TSS with ≥10 normalized reads in at least one of the cell cycle time points were kept for other categories . Some TSSs categorized as N were actually upstream promoters for genes with unusually long 5′ UTRs . All TSSs categorized as N were manually curated to ensure accuracy . On the cell cycle expression profile of each TSS , Fourier coefficients ( coeff 0-7 ) were calculated using normalized reads obtained from synchronized cells . The maximum and minimum normalized read values ( defined as a and b respectively ) from time 0 min to time 140 min were also determined . Cell cycle-regulated TSS are those that meet the following three criteria: 1 . ) coeff1/ ( coeff1+ coeff2 + coeff3 + coeff4 ) ≥ 0 . 35 2 . ) ln ( a ) - ln ( b ) ≥ 1 . 1 3 . ) a ≥ 20 . Regulatory binding sites were identified from groups of promoter sequences upstream of temporally clustered TSSs using MEME [79] and the position of the 5′ nucleotide of the conserved motifs are numbered relative to the TSS sites . The RpoD motif was obtained by searching within 50 bp upstream of constitutively active TSSs . Cell cycle-regulated TSSs were clustered ( k-means , 15 clusters ) by normalized cell cycle profile where the maximum of normalized read value within time 0–140 min is equal to 1 . From the resulting clusters , up to 100 bp of upstream sequences were used to search for enriched CtrA , SciP , SigT , CcrM , and RpoN motifs using MEME . DnaA binding sites from [10] were used to generate a position weight matrix ( PWM ) to search for DnaA motifs within 100 bp upstream of cell cycle-regulated TSSs using FIMO ( p-value setting: <0 . 001 ) [80] . To search for GcrA binding , we searched for ≥3 fold ChIP-seq enrichment within 50 bp upstream of cell cycle regulated TSSs , using ChIP-seq data from [26] . The enrichment was calculated relative to the average coverage across the genome . 75 bp of the upstream sequence of 36 TSSs including primary ( P ) , anti-sense ( A ) , and intergenic non-coding ( N ) TSSs ( see S3 Dataset ) were cloned into the Bgl II and Xho I restriction sites of vector pNJH185 [81] , resulting in transcriptional fusions with the lacZ reporter gene . For the ctrA antisense promoter , 200 bp of upstream sequence was cloned between the Bgl II and XhoI sites . For promoter constructs mutating the SciP binding motif , 105 bp of upstream DNA were cloned between the Bgl II and Xho I sites ( S9 Fig . ) . In each other case , the Bgl II site is upstream and the Xho I site is downstream of the +5 site of the RNA transcript . These 36 constructs and a pNJH185 empty vector control were introduced into Caulobacter crescentus NA1000 cells by electroporation . The LacZ activity of all constructs was measured using mid-log phase NA1000 cell cultures grown in minimal media and according to standard ONPG based β-galactosidase assays . Results in S3 Dataset represent the average of three independent measurements for each strain .
|
The generation of diverse cell types occurs through two fundamental processes; asymmetric cell division and cell differentiation . Cells progress through these developmental changes guided by complex and layered genetic programs that lead to differential expression of the genome . To explore how a genetic program directs cell cycle progression , we examined the global activity of promoters at distinct stages of the cell cycle of the bacterium Caulobacter crescentus , that undergoes cellular differentiation and divides asymmetrically at each cell division . We found that approximately 21% of transcription start sites are cell cycle-regulated , driving the transcription of both mRNAs and non-coding and antisense RNAs . In addition , 102 cell cycle-regulated genes are transcribed from multiple promoters , allowing multiple regulatory inputs to control the logic of gene activation . We found combinatorial control by the five master transcription regulators that provide the core regulation for the genetic circuitry controlling the cell cycle . Much of this combinatorial control appears to be directed at refinement of temporal expression of various genes over the cell cycle , and at tighter control of asymmetric gene expression between the swarmer and stalked daughter cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetic",
"networks",
"caulobacter",
"gene",
"regulation",
"microbiology",
"next-generation",
"sequencing",
"developmental",
"biology",
"genome",
"analysis",
"microbial",
"growth",
"and",
"development",
"bacteria",
"microbial",
"genomics",
"bacterial",
"genomics",
"genome",
"complexity",
"bacterial",
"genomes",
"caulobacter",
"crescentus",
"gene",
"expression",
"systems",
"biology",
"transcriptome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"computational",
"biology",
"organisms"
] |
2015
|
The Global Regulatory Architecture of Transcription during the Caulobacter Cell Cycle
|
ZAP–70 ( Zeta-chain-associated protein kinase 70 ) is a tyrosine kinase that interacts directly with the activated T-cell receptor to transduce downstream signals , and is hence a major player in the regulation of the adaptive immune response . Dysfunction of ZAP–70 causes selective T cell deficiency that in turn results in persistent infections . ZAP–70 is activated by a variety of signals including phosphorylation of the kinase domain ( KD ) , and binding of its regulatory tandem Src homology 2 ( SH2 ) domains to the T cell receptor . The present study investigates molecular mechanisms of activation and inhibition of ZAP–70 via atomically detailed molecular dynamics simulation approaches . We report microsecond timescale simulations of five distinct states of the ZAP–70 KD , comprising apo , inhibited and three phosphorylated variants . Extensive analysis of local flexibility and correlated motions reveal crucial transitions between the states , thus elucidating crucial steps in the activation mechanism of the ZAP–70 KD . Furthermore , we rationalize previously observed staurosporine-bound crystal structures , suggesting that whilst the KD superficially resembles an “active-like” conformation , the inhibitor modulates the underlying protein dynamics and restricts it in a compact , rigid state inaccessible to ligands or cofactors . Finally , our analysis reveals a novel , potentially druggable pocket in close proximity to the activation loop of the kinase , and we subsequently use its structure in fragment-based virtual screening to develop a pharmacophore model . The pocket is distinct from classical type I or type II kinase pockets , and its discovery offers promise in future design of specific kinase inhibitors , whilst mutations in residues associated with this pocket are implicated in immunodeficiency in humans .
ZAP–70 is part of the Syk family of protein kinases , and a key player in the adaptive immune system . [1] It is expressed in T cells and natural killer cells [2] and is essential for their development and function . Inactivating mutations of ZAP–70 cause selective T cell deficiency in humans , which in turn leads to conditions such as severe combined immunodeficiency ( SCID ) [3] and persistent infections [4] . Stimulation of the T cell antigen receptor leads to phosphorylation of tyrosines on intracellular ITAM sequences , favoring recruitment of ZAP–70 via its tandem SH2 domains , and biasing ZAP–70 towards non-auto-inhibited states [5] . Subsequent phosphorylation and auto-phosphorylation events lead to up-regulation of ZAP–70 kinase . In particular , phosphorylation of ZAP–70 at specific tyrosine sites in its KD [6] is likely responsible for its full activation [7] . Mutations of residues Y492 and Y493 in the activation loop of the ZAP–70 KD reveal distinct results for the adjacent phosphorylation sites . The Y492F mutation increases ZAP–70 activity , suggesting that phosphorylation of Y492 is not necessary for the biological function of ZAP–70 . However , the Y493F mutation impairs kinase activation . [8] Downstream targets phosphorylated by ZAP–70 comprise SH2 domain containing leukocyte protein ( SLP–76 ) [9] and an integral membrane protein , linker for activation of T cells ( LAT ) [10] . Both protein targets are crucial for T cell receptor function , and are responsible for downstream signaling and gene transcription . Available three-dimensional structures for the ZAP–70 KD include the isolated domain in complex with staurosporine ( PDB 1U59 [11] ) , a well-characterized ATP-competitive inhibitor of kinases [11–12] . Moreover , structures of the full-length complex of ZAP–70 are available , with the KD bound to ANP , but auto-inhibited by its tandem of SH2 domains . The first such structure ( PDB 2OZO [13] ) included mutations which masked an inhibitory interface between regulatory domain and KD resolved in a subsequent , otherwise similar wild-type structure ( PDB 4K2R [14] ) . Fig 1 illustrates the architecture of the isolated , inhibited KD and highlights significant functional regions . The domain exhibits the distinct bilobal architecture common to other protein kinases , with the activation loop containing Y492 and Y493 located between the two lobes . Staurosporine occupies the ATP binding pocket , which is located at the linkage region between lobes . Despite being bound to inhibitor , Jin et al . reported that the KD is in an active-like state , due to the conformation of the activation loop resembling the geometry of active states observed in the Syk kinase family . However , they noted that the activation loop forms a crystal contact in their structure . Hence , it is unclear whether the loop in the isolated staurosporine complex would likewise adopt an active conformation . Similar conformations of the non-phosphorylated activation loop have been observed for Chk1-staurosporine complexes , although these also involved crystal contacts ( PDB 1NVR [15] ) . The salt bridge formed between K369 and E386 ( residue numbers for ZAP–70 ) , located in the αC helix comprising residues D379 to Q392 , is a conserved motif of active kinase conformations [16] that is normally broken in inactive kinase states , but was formed in the staurosporine ZAP–70 complex . Deindl et al . observed striking similarities between auto-inhibited ZAP–70 and inhibited Hck [17] and c-Src kinase [18] structures . In comparison to the staurosporine complex , the auto-inhibited , ANP-bound crystal structures of ZAP–70 revealed that the αC helix is displaced outwards leading to a loss of this key salt bridge . Available crystal structures presently offer an inconclusive picture of the relation of the ZAP–70 KD conformation to the various intermediate mechanistic states . As observed by Jin et al . , the staurosporine-bound ZAP–70 KD surprisingly appears to adopt an active-like state when compared to other members of the Syk kinase family . Moreover , experimental data from Chan et al . suggests that Y493 phosphorylation is important for catalytic activity of ZAP–70 whereas the neighboring Y492 is not . Motivated by these observations , we have investigated the activation and inhibition mechanisms of ZAP–70 by using molecular simulations of its KD in a number of mechanistic states , over the microsecond timescales necessary to observe key conformational transitions . We discern structural and dynamic properties that yield a molecular basis for previous biophysical and structural experiments , allowing us to rationalize the inhibitor-bound , superficially “active-like” conformation of ZAP–70 , and to propose a scheme for the activation cascade by tracing the phosphorylation-dependent effects through the ZAP–70 KD . Finally , we identify a novel , spontaneously formed cryptic pocket restricted to the non-phosphorylated inactive state of the KD , and use this in virtual fragment-based screening to build a pharmacophore model . The pocket is distinct from classical type I or type II kinase pockets , and hence offers promise in future design of specific kinase inhibitors .
The staurosporine-bound complex remained stable throughout the course of the simulation . This is illustrated by the constant Cα-RMSD ( Fig 2a ) , which rapidly plateaued at ~2–3 Å , as well as the low B-factors ( Fig 2b ) across the entire domain and associated variance throughout the trajectory ( Fig 2c ) . Residual flexibility was observed within the DFG motif , the activation loop downstream of the unmodified phosphorylation sites Y492 and Y493 and the solvent-exposed regions of the αC helix . The DFG motif was observed to form a stable , closed loop structure through a hydrogen bond from the side chain carboxylate-oxygen on D479 through the backbone amide hydrogen on G481 ( Fig 3a ) . This closed loop structure is also present in the staurosporine X-ray structure , 1U59 . The distance of the αC helix from the center of mass of the C-lobe initially increased from a value of 14 Å ( in the crystallographic state ) to 15 Å within ~150 ns and remained stable at this level for the remainder of the trajectory ( Fig 3b ) . Moreover , the phenyl ring of F480 of the DFG motif forms a close contact with the backbone amide of M390 within the αC helix ( S1 Fig ) . The salt bridge K369-E386 was nevertheless present during the entire course of the simulation ( Fig 3c ) . It is present in the staurosporine complex structure 1U59 but absent in the ANP complexes 2OZO and 4K2R . Structural features specific for the staurosporine complex during our simulations comprise an additional salt bridge ( D379-R496 ) formed across the catalytic cleft , and an adjacent intermittent hydrogen bond between the S497 hydroxyl and the N348 side chain amide group . This interaction is not observed in any X-ray structures of ZAP–70 . Moreover , this behavior was not observed in any other complex . These interactions span the catalytic cleft and connect the N- and C-lobes , thereby significantly restricting substrate access . Moreover , these stabilize the position of the activation loop . These structural features are illustrated in Fig 4a . Additionally , an intra-strand interaction hydrogen bond between the side chains of R514 and Y493 could be discerned for significant parts of the trajectories even in the absence of Y493 phosphorylation . Non-phosphorylated ATP-bound ZAP–70 exhibited a similar pattern of flexibility to the staurosporine-inhibited complex . It remained stable and rigid , as illustrated by an RMSD that plateaus similarly to the STA system , and low B-factors ( Fig 2 ) . Both absolute values and variability of B-factors were elevated in the N-lobal regions around residues 500–519 and 537–569 ( Fig 1 , green ) compared to the staurosporine complex ( Fig 2c ) . Unlike the STA system , the non-phosphorylated structure did not present a closed loop within the DFG motif during simulation . The distance from the D479 carboxylate carbon to the backbone amide hydrogen of G481 was stable at ~5 Å during the entire course of the simulation ( Fig 3a ) . The distance between the center of mass of the αC helix and the center of mass of the C-lobe increased from 14 Å to 14 . 5 Å within the first ~100 ns and subsequently remained constant for the duration of the simulation . This behavior is analogous to the behavior of the staurosporine complex ( Fig 3b ) . The contact of F480 of the DFG motif with M390 in the αC helix was significantly less pronounced than in the STA complex or any phosphorylated state ( S1 Fig ) . The salt bridge K369-E386 was present intermittently , around 50% of the total simulation time ( Fig 3c ) . Fig 5 shows Normal modes and projections of the four lowest-frequency modes during the entire trajectory . Mode number 4 was specific for the non-phosphorylated complex , and is associated with a movement of the αC helix towards the remainder of the C-lobe . Concomitantly , the section of the activation loop containing phosphorylation targets Y492 and Y493 moves towards the catalytic cleft , thereby restricting access . ( Fig 5d ) Notably , scanning the surfaces of the KD in the non-phosphorylated complex over 1 . 5 μs of simulation revealed the spontaneous formation of a cryptic pocket adjacent to the activation loop ( Fig 4b ) . This cryptic pocket repeatedly opened and closed from ~700 ns , and reached a maximum volume of ~1400 Å3 ( Fig 4c ) . The protein backbone geometry of the maximum open states encountered at 1050 ns , 1257 ns and 1378 ns is identical . The formation of the pocket primarily arose from the sidechain movement of a single residue , W505 , which is highly conserved across kinase domains . In the initial structure , W505 forms the core of a hydrophobic cluster; thus , its aromatic ring is wedged between P539 and P502 , and is in van der Waal’s contact with V527 , A463 , and the alkyl groups of K504 and R465 . Coupled to the motion of the activation loop , the W505 ring underwent conformational switching within the hydrophobic core ( S2 and S3 Figs ) , with a gradual shift in its position relative to the nearby ATP site , increasing the distance of separation by up to 8 Å over the final ~700 ns ( S3 Fig ) . In its final sidechain orientation ( S3 Fig ) which resulted in the formation of the fully open cryptic pocket , W505 came to rest on the surface of Y506 , W523 , I552 , and W576 , encompassing residues in or nearby to the mobile N-lobal region , whilst remaining in contact with V527 and P502 . The cryptic pocket is primarily formed by residues R460 , D461 , L462 , A463 , K500 , W501 , P502 , W505 , Y506 and S524 , as outlined in Fig 4d . In order to characterize the pocket in more detail , we flexibly docked a commercial library of ~1 , 400 fragments spanning a molecular weight range from 150–300 into this pocket when in its most expanded states . This allowed us to identify the preference of specific regions for defined structural motifs . Following multiple rounds of fragment docking and energy minimization , we used the 100 highest scoring pooled fragments to establish a consensus pharmacophore based on commonly observed features , as depicted in Fig 4d . The five top scoring poses observed across cryptic pocket conformations are illustrated in S4 Fig . The resulting consensus pharmacophore shows an acceptor , a donor , a ring and a hydrophobic feature located within the pocket . Mutational studies previously indicated that Y492 phosphorylation might not be important in ZAP–70 biological function . [8] Simulations revealed that the tyrosine-phosphorylated variant YPY0 exhibited elevated flexibility across the entire protein compared to both the staurosporine and non-phosphorylated ATP complexes . The Cα-RMSD increased gradually over the first microsecond , before plateauing at ~4 Å . Consistently , the B-factor values and their variability were higher in several regions of the protein compared to the non-phosphorylated complexes ( Fig 2 ) . The DFG motif began in a conformation associated with a distance of 5 Å between the D479 carboxylate carbon and G481 backbone amide hydrogen . In contrast to the non-phosphorylated variant , this distance rapidly increased at the ~600 ns mark to 8 Å , indicating a disintegration of the DFG structure ( Fig 3a ) , consistent with geometries observed in X-ray structures of the ANP complexes 2OZO and 4K2R . The αC helix in the YPY0 complex initially adopts a close conformation at a distance of 13 . 5 Å from the center of mass of the C-lobe . After around 500 ns , it underwent a rapid shift , increasing to a distance of 14 . 5 Å as observed in the inhibited and non-phosphorylated complexes ( Fig 3b ) . The K369-E386 salt bridge was absent for the majority of the simulation ( Fig 3c ) . Motions captured in normal modes 1 and 3 are of interest for the YPY0 complex . Normal mode 1 is characteristic for mono-phosphorylation . The mode consists predominantly of an outward movement of the phosphorylated region of the activation loop and a torsional motion of the C-lobe against the N-lobe . ( Fig 5a ) Mode 3 is characteristic of Y492 phosphorylation only . It reveals extension of the activation loop but no associated movement of the αC helix . ( Fig 5c ) Phosphorylation on Y493 is believed to be a decisive activating factor for the KD , in its biological context . It was also observed to show elevated flexibility over the inhibited and non-phosphorylated complexes during simulation . Flexibility was higher than YPY0 variant in terms of RMSD , B-factors and B-factor variability , caused primarily by increased flexibility of the activation loop ( Fig 2 ) . The structure of the equilibrated DFG motif was similar to that observed in the non-phosphorylated variant . The distance between D479 carboxylate carbon and G481 backbone amide hydrogen was constant at approximately 5 Å throughout the simulation ( Fig 3a ) . In contrast to the previously discussed complexes , the αC helix in the Y0YP state adopted a distinct open conformation , indicated by an increased distance from the center of mass of the C-lobe to ~15 . 5 Å within the first ~100 ns ( Fig 3b ) . Consistent with this , the salt bridge K369-E386 was absent for the duration of the simulation ( Fig 3c ) . Strikingly , the phosphorylated Y493 residue did not interact with R514 , despite the complementarity of charge . This peculiar absence of charge-charge interactions extended to the YPYP state . Normal modes 1 and 2 are associated with Y0YP . Whereas normal mode 1 is characteristic for either mono-phosphorylated state , normal mode 2 is specific for Y0YP . Normal mode 2 consists of an extension of the activation loop and concurrent rearrangement of the αC helix . Strikingly , a general opening motion at the hinge region likely allows for easier substrate access to the catalytic cleft . ( Fig 5b ) The di-phosphorylated complex is characterized by the highest flexibility of all investigated states , slightly higher than Y0YP . Cα RMSD indicates high conformational variability ( Fig 2a ) . Analysis of the B-factors profile and associated variability revealed particularly pronounced dynamics within the activation loop and the N-lobal region comprising residues 537–569 ( Fig 2b and 2c ) . This region is associated with the opening of the cryptic pocket observed in the non-phosphorylated Y0Y0 state . In the present YPYP state , however , the segment 537–569 moves independently of the C-terminal activation loop segment , thus not forming the cryptic pocket observed in the Y0Y0 state . The DFG motif was observed to rapidly lose its structure , and spontaneously adopted an extended conformation similar to the YPY0 mono-phosphorylated state , with a D479 carboxylate carbon to G481 backbone amide hydrogen distance of 8 Å after approximately 50 ns . This extended conformation was present for the duration of the simulation ( Fig 3a ) and can also be observed in the 2OZO and 4K2R ANP-bound crystal structures . The conformation of the αC helix observed in the present state was analogous to the conformation in the Y0YP state , at a distance of 15 . 5 Å from the C-lobe center of mass ( Fig 3b ) , whilst the salt bridge K369-E386 was not present for the duration of the trajectory . Di-phosphorylation was not represented in a singular normal mode . Generally , YPYP exhibited a combination of structural and dynamic features observed individually for the YPY0 and Y0YP states .
Common patterns observed in all simulated states were characterized by two distinct , comparatively rigid cores of the C-lobe and N-lobe , as well as flexible segments of the activation loop and the region formed by residues 537–569 . Baseline flexibility in the non-phosphorylated and inhibited states was significantly lower than in either mono- or di-phosphorylated states , as indicated by the calculated B-factors . In order to differentiate between the alternative phosphorylated states , it should be noted that experimentally Y492F does not adversely affect ZAP–70 activity , whereas Y493F abolishes ZAP–70 function . [8] Therefore , we can use the effects of Y492 phosphorylation as a baseline for dynamic changes induced by monophosphorylation at the C-terminal end of the activation loop and contrast its effects with those observed in states containing phosphorylated Y493 . The differences in dynamics between the two states offer some indication of the functional relevance of the observed changes across systems . Whereas phosphorylation caused a global increase in protein flexibility , specific changes were localized around the activation loop as well as the αC helix region . Strikingly , the conserved DFG motif at the N-terminal side of the activation loop adopted three distinct states across the different simulation systems , characterized by a cyclization through hydrogen bonding . By considering the distance between the carboxylate carbon of D479 and the backbone amide hydrogen of G481 , we could identify a cyclic , closed state at approximately 3 Å , a semi-closed state at a distance of ~6 Å , and an open state at a distance of ~8 Å ( Fig 6 ) . We surmise that the semi-closed state of the DFG motif is relevant for the catalytic activity in ZAP–70 as it occurs in the un-phosphorylated state as well as in the Y0YP state . YPYP and YPY0 bias the DFG conformation towards the open state whereas staurosporine inhibition results in stabilization of the closed state . We postulate that YPY0 causes a repulsive charge-charge interaction with D479 , thereby promoting the open state . As staurosporine is bereft of a negative charge proximal to the DFG motif and misses an Mg2+ ion , D479 is free to position itself in a hydrogen bonding orientation towards the backbone N-H of G481 . This behavior of the DFG motif is consistent with differing orientations of D479 observed in the staurosporine-bound X-ray structure 1U59 versus the ANP-bound complexes 2OZO and 4K2R . Generally , our observations signify that residue D479 located close to the catalytic center is highly sensitive to its local electrostatic environment . It should be noted that all simulations started from the staurosporine-bound protein conformation represented by 1U59 , as it is the only structure in which all residues of the activation loop are resolved . While the choice of this starting state may introduce a bias towards active-like states in the remaining simulations , the reorientation of the DFG motif is consistent with the 2OZO and 4K2R structures . Thus we assume that the simulation times are sufficient to sample at a minimum conformational transitions between the active and intermediate states [19] . Despite the length of our trajectories , we were unable to observe DFG-in/DFG-out transitions as indicated by the interaction of F480 with M390 in the αC helix ( S1 Fig ) . A recurring motif in Src kinase family activation patterns is the formation of a salt bridge from the bulk of the C-lobe to the αC helix . In ZAP–70 this salt bridge may form between R369 in the C-lobe and E386 in the αC helix . This connection is associated with motions of the αC helix that have previously been identified as crucial in Src kinase family active states . Intermittent closing and opening of this salt bridge was observed throughout the course of the simulation for the non-phosphorylated , ATP-bound kinase , thus indicating a fine energetic equilibrium at this state point . Inhibition by staurosporine caused this salt bridge to adopt a permanently closed position , consistent with that observed in the 1U59 crystal structure . All phosphorylated states revealed an increased average distance between the guadinium group of R369 and the E386 carboxylate function , as well as elevated variance of this distance compared to the non-phosphorylated state . Positioning of the αC helix relative to the bulk of the protein fits with this pattern . Inhibited and non-phosphorylated states show a closer positioning to the protein center of mass of the αC helix . Phosphorylation of Y492 led to a similar position of the αC helix as the inhibited and non-phosphorylated states . Only phosphorylation of Y493 or di-phosphorylation induced repositioning of the αC helix , consistent with kinase activation . From these observations we conclude that the strength of the salt bridge R369-E386 is weakened by phosphorylation and is a necessary step for subsequent repositioning of the αC helix towards an active conformation . This repositioning is solely observed if Y493 is phosphorylated . Mono-phosphorylation of Y492 weakens the salt bridge , but does not induce a conformational shift in the αC helix . Inhibition of ZAP–70 was observed to be associated with a general reduction in flexibility across the KD . Moreover , all investigated structural features comprising the DFG motif , the αC helix , the R369-E386 salt bridge , as well as patterns of normal modes and B-factors , indicate lowered dynamics for the staurosporine complex . Our analyses suggest that binding of staurosporine traps the ZAP–70 KD in a compact and rigid state . The observed rigidity would limit both substrate access and the likelihood of competitive binding by co-factor at the catalytic center . Our analyses have allowed us to identify a novel pocket adjacent to the active site and the activation loop . This pocket is neither a classical type I or type II kinase pocket and offers potential possibilities for design of new specific inhibitory ligands . Interestingly , the pocket only occurs in the non-phosphorylated state . Whereas we did not expect to find it in the very compact and rigid STA complex , it is quite surprising that the pocket was not observed in the generally more flexible phosphorylated states . Closer examination of the normal modes reveals that the phosphorylated states exhibit concerted motions along the entire activation loop , whereas these motions are much more localized to the region following the two tyrosines Y492 and Y493 in the non-phosphorylated state . We therefore postulate that the opening of this pocket is facilitated by a flap-like rearrangement of residues 495–498 , associated with the “gating” of the conserved residue W505 , which switches to an alternative hydrophobic environment supported by residues in or nearby to the mobile N-lobal region . Strikingly , of seven known loss-of-function mutants in the ZAP–70 kinase domain reported to lead to SCID in humans [3] , five are either associated with residues that initially contact W505 or reorient and interact with it during the gating process . These include R465 ( two reported missense mutants ) , part of the highly conserved DLAARN motif , and K538 ( 13 base pair deletion ) in the flexible loop region , both of whose alkyl groups form van der Waal’s interactions with W505 , along with A507 and S518 ( missense mutants ) which interact with W505 and reorient during gating and cavity formation . The remaining two , M572 ( missense ) and K541 ( splicing error ) , are expected to change the local environment of W505 in its final , gated state . Finally , two hypomorphic ZAP–70 kinase mutants in mice with partial defects in TCR signaling included mutation to arginine of W504 ( equivalent to W505 in humans ) alone or in combination with I367F [20] . Thus we surmise that drugging this cryptic pocket could prove valuable in the study of SCID . Proceeding work will focus on targeting this pocket with small-molecules and establishing the validity of our approach by experimentally probing the chemical and structural biology of this site . We presume that increasing the size of residues lining this pocket through mutational modifications F516W , D521E or S524T could bias the system towards a stable , permanently open conformation of the cryptic pocket . Phosphorylation was observed to lead to a global increase in dynamics , irrespective of the precise phosphorylation state . Double phosphorylation at Y492 and Y493 had a significantly stronger mobilizing effect than single phosphorylation of either residue . However , structural changes were distinctly different for the individual mono-phosphorylated complexes . Mutational studies suggest that Y492 is only weakly implicated in biological activation of ZAP–70 . However , Y493 phosphorylation is of crucial importance to biological function as the Y493F mutation abolishes catalytic activity . Our simulations allow us to identify changes that are specific to Y493 phosphorylation and therefore let us trace the activation cascade: Y493 phosphorylation causes the salt bridge R369-E386 connecting the C-lobe with the αC helix to weaken . In contrast to Y492 phosphorylation , Y493 phosphorylation also promotes rearrangement of the αC helix towards an active conformation already observed for members of the Src kinase family . Concurrently , flexibility of the activation loop increases significantly , thus allowing for easier access to the catalytic center . A similar increase in activation loop exposure has been observed in the activation of focal adhesion kinase ( FAK ) . However , the activation cascade of FAK is not directly analogous to ZAP–70 , as FAK has an additional FERM domain involved in forming an auto-inhibited complex . [21] Our proposed activation cascade for ZAP–70 is summarized in Fig 7 . Normal mode analysis further confirms these motions as characteristic for the biologically relevant Y493 phosphorylation . In the present study , we investigated the mechanisms underlying ZAP–70 activation and inhibition . We were able to identify crucial motions associated with biological activation . We demonstrated how subtle changes in these patterns trap the enzyme in an inhibited state that superficially resembles an “active-like” structure , and derived a viable microscopic mechanism of ZAP–70 KD activation by phosphorylation . Furthermore , our simulations have allowed us to identify a cryptic pocket that is neither a classical type I nor type II binding site , and may offer promise as a new site for specific targeting by small molecule ligands .
Initial coordinates for ZAP–70 KD were obtained from the PDB [22] structure 1U59 . [11] All histidine residues were set to be deprotonated and in the ε-NH tautomeric state . All other ionisable residues were set in their default , charged state . No cysteine residue is involved in disulfide bond formation . This structural template was used to construct five distinct complexes . These comprised one complex with the bound staurosporine inhibitor , and Y492 and Y493 in their non-phosphorylated states ( STA ) . Moreover , four ATP-bound complexes were built in various phosphorylation states , namely: non-phosphorylated at Y492 and Y493 ( Y0Y0 ) , Y492 phosphorylated ( YPY0 ) , Y493 phosphorylated ( Y0YP ) , and both Y492 and Y493 phosphorylated ( YPYP ) . All ATP-bound complexes were constructed by superimposing ATP and the Mg2+ ion from PDB 4K2R onto PDB 1U59 . Parameters for staurosporine were described by CGenFF version 2b5 . [23] Protein interactions were modeled using the CHARMM22/CMAP force field . [24] [25] All systems were solvated in a 0 . 1 M sodium chloride solution containing approximately 12 , 000 TIP3P [26] water molecules . Solvation resulted in rectangular box sizes of approximately 8 . 4 x 8 . 4 x 5 . 9 nm . The systems were equilibrated by performing 1000 steps of steepest descent minimization followed by a series of three 500 ps of NpT ensemble simulations with gradually decreasing position-restraints on the protein and ligand heavy atoms . All simulations were performed using GROMACS 4 . 5 . 5 . [27] Electrostatic interactions were described using particle mesh Ewald . [28] Van-der-Waals and Ewald cut-offs were set to 1 . 2 nm . Bonds to hydrogen atoms were constrained with the LINCS algorithm [29] allowing an integration time step of 2 fs . Temperature was controlled for distinct coupling groups of solvent and solute using separate v-rescale thermostats [30] at 298 K , using a coupling constant τ of 1 ps . An isotropic Parrinello-Rahman barostat [31] maintained a pressure of 1 atm , using a coupling constant τ of 5 ps . Following equilibration , all systems were simulated for 1 . 5 microseconds in the NpT ensemble . Frames were saved every 100 ps yielding a single , continuous trajectory of 15 , 000 frames for each system . Analysis of the trajectories was performed using cpptraj from the AmberTools 14 package . [32] All trajectories were aligned by their Cα atoms . Subsequently , backbone RMSD and B-factors were calculated as illustrated in Fig 2 . Individual trajectories were split into 10 sequential parts of equal length , and B-factors were evaluated separately for each part . This allowed us to assess convergence , by identifying whether local flexibility and the average structure remains constant throughout the simulation or is subject to change . Regions of particular interest comprise the activation loop , the conserved DFG motif containing D479 , F480 and G481 , [33] and the αC helix with the associated salt bridge K369-E386 . Relative positions and conformations of these regions as a function of simulation time are summarized in Fig 3 . Subsequently , trajectories of all systems were concatenated and normal modes were calculated . [34] Projections were obtained and the amplitudes of these projections were correlated with the respective states . This procedure allowed us to identify crucial conformational transitions associated with each state . Normal mode components and projections for the three lowest-frequency non-zero normal modes are given in Fig 5 . trj_cavity [35] was used to screen for novel transient pockets occurring during the course of the simulation . The three frames of the non-phosphorylated trajectory that show the largest volume for the cryptic pocket , at 1050 ns , 1257 ns , and 1378 ns respectively , were selected for fragment screening . These frames were imported into Schrodinger Maestro 2015–2 . Subsequently , the prepwiz tool was used to minimize hydrogen atoms while keeping protein heavy atoms restrained . A commercial library of 1430 fragments spanning a molecular weight range from 150–300 were prepared using the ligprep tool to generate protomers and tautomers . A receptor grid centered on the pocket residues R460 , D461 , L462 , A463 , K500 , W501 , P502 , W505 , Y506 and S524 was generated for each of the three structures . The ligand library was docked using GlideSP . Ligands were allowed to be flexible during docking . Five initial states per ligand were retained , minimization was carried out and the highest-ranking pose after minimization was retained . The 100 highest scoring fragments from each structure were pooled and used as a basis to create a pharmacophore model of the cryptic pocket using Prime . Settings required a hypothesis of at least 50 ligands matching 3 to 5 features . We selected the pharmacophore with the highest number of fitting poses as the consensus pharmacophore .
|
ZAP–70 is a key protein kinase in the adaptive immune system . It is essential for development and function of T cells and natural killer cells , and associated mutations can lead to conditions such as severe combined immunodeficiency ( SCID ) . Here , simulations of the ZAP–70 kinase domain are used to study its dynamics in response to different mechanistic signals . We identify crucial motions over microsecond timescales , which help to rationalize in atomic detail previous structural and experimental data regarding its biological regulation . We subsequently propose a scheme for the phosphorylation-dependent activation cascade of ZAP–70 , and for its ligand-dependent inhibition . Finally , we characterize a novel cryptic pocket adjacent to the active site and activation loop , which is distinct from classical type I or type II kinase sites . The pocket is in close proximity to several residues whose mutations cause SCID in humans , and its identification offers promise in future drug design efforts .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
The Structural Basis for Activation and Inhibition of ZAP-70 Kinase Domain
|
Protein–protein interactions are challenging targets for modulation by small molecules . Here , we propose an approach that harnesses the increasing structural coverage of protein complexes to identify small molecules that may target protein interactions . Specifically , we identify ligand and protein binding sites that overlap upon alignment of homologous proteins . Of the 2 , 619 protein structure families observed to bind proteins , 1 , 028 also bind small molecules ( 250–1000 Da ) , and 197 exhibit a statistically significant ( p<0 . 01 ) overlap between ligand and protein binding positions . These “bi-functional positions” , which bind both ligands and proteins , are particularly enriched in tyrosine and tryptophan residues , similar to “energetic hotspots” described previously , and are significantly less conserved than mono-functional and solvent exposed positions . Homology transfer identifies ligands whose binding sites overlap at least 20% of the protein interface for 35% of domain–domain and 45% of domain–peptide mediated interactions . The analysis recovered known small-molecule modulators of protein interactions as well as predicted new interaction targets based on the sequence similarity of ligand binding sites . We illustrate the predictive utility of the method by suggesting structural mechanisms for the effects of sanglifehrin A on HIV virion production , bepridil on the cellular entry of anthrax edema factor , and fusicoccin on vertebrate developmental pathways . The results , available at http://pibase . janelia . org , represent a comprehensive collection of structurally characterized modulators of protein interactions , and suggest that homologous structures are a useful resource for the rational design of interaction modulators .
Protein–protein interactions are a broad class of therapeutic and chemical biology targets [1] . Traditionally these targets were thought to be refractory to small molecule modulation . However , recent efforts have led to interaction modulators that are valuable tools in mapping signalling networks and are entering clinical trials for therapeutic use [2] . Although natural substrates often serve as guides for rational drug design , such information is rarely available for protein–protein interfaces [3] . Here we attempt to provide such a starting point through a structural analysis of known protein and ligand binding sites . We posit that although ligands that are known to bind to specific protein–protein interfaces are rare , examples of ligands that bind to corresponding positions in homologous proteins may be available . These homologous sites , and the ligands they bind , may serve as starting points for rationally designing small molecule modulators of protein interactions . The physicochemical , geometric , and evolutionary properties of ligand and protein binding sites have been extensively studied by analyzing three-dimensional protein structures [4]–[6] . On average , protein interfaces are relatively planar , more physically adaptable , and much larger than the small , rigid , pockets that bind small molecules [5] , [7] . Despite the large total surface area of protein interfaces , a small subset of these residues , termed ‘hotspots’ , contribute disproportionately to the affinity of protein–protein interactions [8]–[10] . Small molecules that target these hotspots have been found to effectively compete against proteins in binding events [11] . The computational methods developed for traditional rational drug design , such as pocket detection and virtual screening , have also been applied to identify small molecules modulators of protein interactions . The methods are frequently adapted to the unique properties of protein interfaces , such as their adaptivity in forming small transient cavities that can bind small molecules [12] . This property led to the use of molecular dynamics simulations to search protein interfaces for transient pockets that are subsequently targeted by virtual screening [13] . In this study , we take a conceptually related approach that harnesses the conformational ( and chemical ) space sampled by homologous members of a protein family . The magnitude and direction of this evolutionary sampling has been found to correlate with the conformational space sampled physically by an individual member of a protein family [14]–[16] . Here , we perform a systematic analysis of structurally characterized ligand and protein binding sites , with a central goal of comprehensively identifying , enumerating , and describing those positions in protein structure families where both ligands and proteins have been observed to bind . We first analyze the overlap of these binding sites within protein families , characterizing the composition and conservation of these ‘bi-functional’ positions , and identifying the families in which they are more or less prevalent than expected by chance . Next , we describe protein–protein and protein–peptide interactions for which small molecules were observed to bind at corresponding or homologous positions in other protein structures . Finally , we describe known interaction modulators recovered by the analysis , and illustrate its predictive utility by suggesting structural mechanisms for the observed effects of three small molecules .
We began by assembling a comprehensive list of protein and ligand binding sites . Protein–protein ( inter-molecular domain–domain , intra-molecular domain–domain , and domain–peptide ) binding sites were obtained from PIBASE ( v200808 ) [17] , based on domain boundaries and classifications from SCOP ( v1 . 73 ) [18] ( details in Materials and Methods ) . Peptide binding sites were included in the analysis because the structures of protein complexes are often solved with only the peptides that mediate the interaction , rather than the full-length protein . Ligand binding sites were obtained from LIGBASE [19] , and mapped onto SCOP domains using family alignments from the ASTRAL compendium [20] . Binding sites that shared more than 90% of their corresponding alignment positions were grouped together and a representative was chosen randomly , yielding a final dataset of 35 , 168 ligand binding sites , 2 , 332 peptide binding sites , 12 , 015 inter-molecular domain interfaces , and 4 , 290 intra-molecular domain interfaces , for all of which the structure is known ( Table S3 ) . This redundancy removal procedure ( Materials and Methods ) partially corrects the human bias in structural coverage of proteins , protein complexes , and protein-ligand complexes . Other aspects of bias can not be corrected and therefore affect our observations; For example , the analysis is limited to those proteins , ligands , and complexes that have been structurally characterized . We first quantified the extent and significance of overlap between all ligand and protein binding sites observed in each protein family . The binding sites were mapped onto alignments of domain families obtained from the ASTRAL compendium [20] ( Fig . S1B ) . This mapping procedure implicitly accounts for redundant structures , as multiple structures of the same binding site do not contribute any additional positions beyond those protein-binding or ligand-binding positions identified by the first structure . Of the 2 , 619 families that bind proteins , 1 , 028 also bind small molecules , and 736 of these have at least 5 bi-functional positions ( Table S1 ) . The overlap of ligand and protein binding sites within each family was quantified using the numbers of alignment positions at which ligands ( ) , proteins ( ) , or both ligands and proteins ( ) were bound , as well as the number of solvent-exposed positions ( ) . ( 1 ) An alignment position was considered solvent-exposed if at least one of the domains in the family had a residue with side-chain solvent exposure of greater than 7% at that position ( MODELLER v9 . 4 [21] ) . The statistical significance ( Fisher's exact one-tailed p-value ) of the observed overlap for each family was assessed against a null model in which the ligand and protein binding site positions are randomly and independently placed at solvent-exposed positions ( R v2 . 5 . 1 , http://r-project . org ) . We identified 197 families with significantly more ( right-tail p-val0 . 01 ) , and 113 families with significantly fewer ( left-tail p-val0 . 01 ) , bi-functional positions than expected by chance ( Fig . 1A , Table S2 ) . These two sets of families exhibit differences in the distribution of functions as defined by SUPERFAMILY [22] ( Fig . S1D ) . The significance of the function propensity values were estimated by a non-parametric bootstrap sampling procedure to compute 95% confidence intervals ( Table S4 , Materials and Methods ) . Families with significantly less overlap ( p-val ) than expected by chance were enriched in Metabolism and depleted in Regulation ( ) . In contrast , families with significantly more overlap ( p-val ) than expected by chance were depleted in Metabolism and enriched in Intracellular processes ( ) . For example , ten of the overlapping families are involved in signal transduction compared to none of the non-overlapping families . We next asked whether the chemical or evolutionary properties of bi-functional positions were different from other positions that were part of only ligand or protein binding sites ( mono-functional ) or solvent-exposed . The propensities of each amino acid residue at mono-functional and bi-functional positions were calculated relative to all exposed residues , and their significance estimated by a bootstrap resampling procedure ( Fig . 1B , Table S5 , Materials and Methods ) . The magnitudes of these propensities are within the range reported in previous binding site analyses [4] , [23] . The propensity of residue types that exist at the bi-functional positions are generally intermediate between those of ligand-only and protein-only positions , although they are more similar to the ligand-only positions ( Fig . 1B ) . In particular , bi-functional positions have a higher propensity of tryptophan , histidine , and phenylalanine residues relative to both protein-binding positions and solvent exposed residues . In addition , bi-functional positions have a higher propensity for tyrosine , and slightly lower propensities for alanine , isoleucine , leucine , and valine , than either mono-functional or solvent-exposed positions . Bi-functional positions are also significantly less conserved than mono-functional or solvent exposed positions , as measured by an entropy-based conservation score ( Fig . 1C ) as well as a simple count of residue types ( Fig . S1E ) . This lower conservation was considered statistically significant ( p-val ) by both Kolmogorov-Smirnov and Mann-Whitney tests ( Materials and Methods ) . Although it is difficult to precisely identify the reason for the lower conservation of bi-functional positions , one possible explanation is related to the definition of these positions . We identified bi-functional positions because they participate in different functions – ligand binding and protein binding – in different family members . These different functions might require different residue type compositions , resulting in a lower conservation score for these positions . We also observed minimal , although statistically significant ( p-val ) , differences in conservation between mono-functional and solvent-exposed residues: ligand-only positions were more conserved than all exposed residues , which in turn were more conserved than protein-only positions . The small magnitude of the difference in conservation between mono-functional and all exposed residues is in agreement with previous findings that conservation alone is of minimal predictive use for the identification of binding sites [6] . Having established that ligand and protein binding sites often overlap within protein families , we aimed to determine the utility of known ligand binding sites for targeting particular protein–protein interactions . The ligand binding sites were mapped onto individual domain–domain and domain–peptide interfaces , using ASTRAL alignments as described earlier ( Fig . S1C ) . The overlap between each ligand binding site and protein interface was characterized by the fraction of interface residues aligned to ligand binding site residues . ( 2 ) When the ligand binding site aligned to both sides of a domain–domain interface , the larger of the two overlap fractions was used as the overlap score . The ligand binding site coverage of each protein–protein interface was summarized using two scores . First , a maximal overlap score was used to quantify the maximum overlap observed by any ligand for the protein–protein interface . Second , a cumulative overlap score was computed by simultaneously aligning all homologous ligand binding sites onto each protein–protein interface and calculating the fraction coverage . This procedure is conceptually related to fragment-based drug discovery techniques , such as tethering [24] . The behavior of these overlap scores was examined as a function of the sequence identity between the ligand binding site and the corresponding positions in the interacting proteins ( Fig . 2 , S2 ) . As expected , the coverage of interfaces was reduced at higher thresholds of sequence identity ( Fig . 2A , 2B ) , and the distributions of cumulative overlap scores ( Fig . S2G , S2H , S2I ) exhibit a higher interface coverage than the corresponding distributions of maximum overlap scores ( Fig . S2A , S2B , S2C ) . In addition , the domain–peptide interfaces have higher binding site overlaps ( Fig . 2B ) , on average , than domain–domain interfaces ( Fig . 2A ) . This observation is likely due to the smaller sizes of domain–peptide interfaces , which are thus more readily covered by small molecule binding sites . Although the analysis suggests that most interfaces do not have a homologous ligand binding site , as seen by the main peak over an interface overlap of 0 ( Fig . 2C ) , there are a significant number of interfaces for which overlapping homologous ligand binding sites do exist . In particular , a significant number of protein interfaces overlap with homologous ligand binding sites of greater than 30% sequence identity , previously determined to be a reliable threshold for homology transfer of ligand binding sites [25] . The systematic alignment of ligand binding sites onto protein interfaces generates a dataset useful for two primary purposes . First , it serves as a comprehensive collection of structurally characterized interaction modulators , in the cases where the ligand binding domain is identical to the sequence involved in the protein interaction ( Table 1 ) . Second , it serves as a set of predicted interaction modulators , where the ligand binding site itself is highly similar to the corresponding region in the target interaction , but the overall domain is only homologous , rather than identical ( Table S6 ) . To validate the accuracy of the mapping method , we checked whether known protein interaction modulators were recovered by the method . Indeed , all but one of the modulators discussed in a recent review article [2] were identified by the method: Interleukin-2 – Interleukin-2 receptor ( PDB 2ERJ:A , D; 1PY2:FRH ) , MDM2–p53 ( 1T4F:M , P; 1T4E:DIZ ) , HPV E2–E1 helicase ( 1TUE:A , B; 1R6N:434 ) , ZipA–FtsZ ( 1F47:A , B; 1Y2F:WAI ) , and TNF- homotrimer ( 2TNF; 2AZ5:307 ) . The interaction between Bcl-X–BAD ( PDB 2BZW ) was missed by our analysis because the ligand bound structure ( 2YXJ:N3C ) was published too recently to be classified in the current SCOP domain database . The nearly complete recovery of known modulators suggests that the binding site data used in the analysis and the procedure used to map them operated correctly . We present additional examples of ligand binding sites that overlap interfaces to demonstrate the diversity of interactions for which ligand binding has been observed ( Table 1 ) . Having established the accuracy of the binding site mapping , we next examined the results for their predictive utility in identifying small molecule modulators of protein interactions . Those ligand binding sites that mapped with a high sequence identity , in the context of different protein sequences , represent high confidence predictions where ligand binding may occur ( Table S6 ) . This kind of prediction is an extension of the widely used homology-transfer concept in protein function annotation [25] . The ligands identified in the analysis fell into four broad categories based on the kinds of protein–protein interactions that they overlapped ( Table 1 , S6 ) . The most frequently observed category were synthetic enzyme inhibitors that overlapped with the interfaces between enzymes and their protein or peptide inhibitors . These interactions include carboxypeptidase , ribonuclease , trypsins , coagulation factors , and their protein inhibitors ( Fig . 3A ) . The high number of ligands identified in this class is not surprising , as enzyme–inhibitor complexes are among the most extensively structurally characterized and targeted by synthetic inhibitors . A related group of ligands overlapped with the interface of an enzyme and its natural protein or peptide substrate . This class includes ligands that bound at signaling complexes such as MDM2–p53 , farnesyltransfrease–h-ras , and histone acetyltransferase–p53 . An example that is used therapeutically are HIV protease inhibitors bound at the protease dimer in place of its peptide substrate ( Fig . 3B ) . We also include enzyme homodimers in this group , such as the transketolase and the ornithine decarboxylase homodimers ( Table S6 ) . A third class of ligands overlapped with the interface of structural or regulatory protein–protein interactions . These ligands include natural toxins , such as kabiramide C bound at the actin–gelsolin interface ( Fig . 3C ) and fusicoccin bound at the interface of 14-3-3 proteins ( Fig . 4B ) . This class also includes synthetic compounds such as ajulemic acid that bound at the interface of peroxisome proliferator activated recpetor gamma ( PPARG ) and the LXXLL coactivator ( Table 1 ) . The fourth group of ligands were transferred from structures where they were present at domain interfaces . Although it is difficult to predict the effect of these ligands on the target interface , this group of ligands may be more likely to sterically complement protein interfaces than ligands in the other groups , which more likely sterically hinder protein interactions . This group includes elaidoylamide bound at the homodimeric interface of agkistrodotoxin Phospholipase A2 ( PDB 1RGB ) , and bepridil bound at the interface of Troponins C and I ( 1LXF; Fig . 3D ) . Ligands in this class may be of potential use for designing chemically induced dimerization systems [26] . This technique relies on the ability of particular small molecules , such as Rapamycin and FK506 , to simultaneously bind two proteins , and has been extensively used to study and control cell signaling processes . This group of ligands also slightly overlaps with the second group , as HIV protease inhibitors bind at the homodimeric protease interface ( Fig . 3B ) . Natural ligands such as ATP , GTP , GNP also often bind at domain interfaces . Another class of protein complexes with overlapping homologous ligand binding sites are antibody–antigen complexes . These overlaps are an expected result of the diversity of the complementary-determining regions of immunoglobulins that enable binding to virtually all proteins and small molecules . The ligands that mapped to intra-molecular domain interfaces included natural ligands such as ATP , GTP , and Heme groups , as well as synthetic and natural toxins such as the Pulvomycin and Kirromycin antibiotics ( Table S7 ) . Since we focus on direct modulators of protein–protein interactions , we will not discuss these ligands . However , ligands that bind at intra-molecular domain interfaces may serve as logical switches in cellular signaling networks [27] . Although we observed overlaps that occur in a variety of functional classes , they can all contribute towards a structural understanding of bi-functional positions . Irrespective of the natural or synthetic source of the small molecule , or the particular functional class of protein interaction , the resulting overlaps are structurally informative for understanding what makes particular interface regions amenable to targeting by small molecules . This point can be further clarified by considering the known modulators of protein interactions that we used to test the fidelity of our mapping procedure . Although these examples involve synthetic small molecules , they have been extensively characterized structurally to understand what makes their particular binding sites amenable to targeting by small molecules [2] . Ignoring these examples because of their synthetic source would discard useful structural information . The results also suggest possible structural mechanisms for the observed effects of small molecules . We will describe three such examples , each from a different ligand class: sanglifehrin A , bepridil , and fusicoccin . Sanglifehrin A is an immuno-suppressant , synthesized by an Actinomycetes species , that has been observed to reduce HIV-1 virion production [28] . Our structural analysis found that its binding site on cyclophilin A [29] overlapped completely with the complex formed by cyclophilin A and the HIV capsid [30] ( Fig . 4A ) . This overlap suggests that sanglifehrin A competes with the HIV protein for interaction with cyclophilin A . This prediction is in agreement with biochemical evidence that describes a reduction in virion production by sanglifehrin A through a cyclophilin-dependent mechanism [28] . Fusicoccin is a toxin , synthesized by the fungus Fusicoccum amygdali , that disrupts protein interactions mediated by plant 14-3-3 proteins [31] . Here we observed that its ligand binding site is nearly conserved in mammalian 14-3-3 proteins and overlaps with the 14-3-3-–Seretonin N-acetyltransferase and 14-3-3-–R18 peptide interfaces ( Table S6 , Fig . 4B ) . This high level of binding site similarity suggests that fusicoccin also modulates animal 14-3-3 interactions . In fact , this modulation has been observed experimentally , with fusicoccin used as a tool to disrupt 14-3-3 interactions involved in early left-right developmental patterning in Xenopus [32] . Bepridil is an FDA-approved calcium channel blocker that was until recently used to treat refractory angina . Recently it was found to inhibit the cellular entry of two anthrax toxin components: the edema and lethal factors [33] . Here we observed that the troponin C binding site for bepridil [34] transfers with high sequence identity to the calmodulin–anthrax edema factor interface [35] . The ASTRAL family alignment transferred the binding site to the first calmodulin EF-hand that is not directly in contact with the edema factor . In this alignment , the binding site overlap is minimal ( 1 of 46 protein interface residues; Table S6 , Fig . 4C ) and occurs at the periphery of the interaction . However , upon visualization , it was found that the second EF-hand also aligns well with troponin C , and in this alignment the bepridil binding site directly overlaps with the edema factor interface ( Fig . 4C ) . This alignment suggests that bepridil may disrupt the calmodulin–edema factor interaction by binding to calmodulin . This hypothesis , based on structural data alone , is in agreement with experimental findings that describe reduction in the lethality of edema factor by bepridil [33] .
We observed that several small molecule compounds , originally designed for traditional medicinal chemistry targets such as enzyme active sites , in fact target protein interfaces . These include several FDA-approved drugs , such as bepridil that binds at the interface between Troponins C and I , and HIV protease inhibitors that bind at the dimer interface . Although these examples involve fairly small protein interfaces , this observation suggests that protein–protein interactions are not completely novel targets for medicinal chemistry , and that the chemical , biophysical , and computational experience that has been developed in traditional rational drug design may also be applicable to interaction targets . As protein interaction networks are resolved with greater accuracy and coverage [47] , small molecules become important perturbation tools to examine their functional significance . In addition , a therapeutic application that is becoming increasingly relevant is the targeting of host–pathogen protein interactions , which have been the subject of recent investigations using high-throughput experimental [48] , [49] and computational [50] , [51] methods . These interactions may be a valuable alternative to traditional targets for the increasingly difficult challenge of antibiotic development [52] , [53] . We expect our results , available in PIBASE ( http://pibase . janelia . org ) , to serve as a structural resource to aid in the rational design of small molecule modulators of protein–protein interactions .
Residues in domain–domain and domain–peptide binding sites were obtained from PIBASE v200808 [17] based on domain boundaries and classifications from SCOP v1 . 73 [18] . Peptides were defined as those chains at least 5 amino acid residues long that were not classified by SCOP or were classified in the “peptide or fragment” SCOP class . Binding sites were defined as residues containing at least one non-hydrogen atom within 5 Å of the interacting domain or peptide . Domain–domain interfaces were filtered using a threshold of at least 500 inter-atomic contacts at a distance threshold of 5Å ( 500 Å2 buried surface area ) , to remove small interfaces that are often crystallographic artifacts . A minimum domain participation of 5 residues was also imposed on domain–peptide interactions to remove small interfaces . This procedure identified 24 , 717 inter-molecular domain–domain , 13 , 228 intra-molecular domain–domain , and 6 , 911 domain–peptide interactions involving 2540 , 1485 , and 534 domain families , respectively . Ligand binding sites were obtained from LIGBASE [19] , defined as residues with at least one non-hydrogen atom within 5Å of the ligand . The analysis was restricted to PDB HETERO groups with molecular weights between 250–1000 Daltons , as this range removes crystallographic buffers and small ions present in many PDB entries , and also encompasses most orally administered drugs . MDL and CIF formatted descriptions of the ligand structures were obtained from the MSD Ligand Chemistry dictionary [54] . This procedure identified 39 , 085 binding sites on domains from 1 , 131 families . Redundant binding sites were identified by single-linkage clustering of binding sites that shared more than 90% of their residues as measured by: ( alignment positions shared by the two binding sites ) / ( positions in either binding site ) . This reduced the number of ligand binding sites from 39 , 085 to 35 , 168; peptide binding sites from 4 , 937 to 2 , 332; inter-molecular domain interfaces from 40 , 791 to 12 , 015 , and intra-molecular domain interfaces from 17 , 863 to 4 , 290 . The redundancy removal was performed with respect to the alignment positions , rather than amino acid sequence identity , because the binding site projection procedure relied on the alignment positions . This redundancy removal procedure aimed to reduce the effect of PDB bias in structural coverage of proteins , protein complexes , and protein-ligand complexes . The propensity of residue types in each class of position ( ligand-only binding , protein-only binding , or bi-functional ) was computed relative to all solvent exposed positions by counting the frequency of the 20 standard amino acid residue types: ( 3 ) Residue types that occur more frequently at a particular binding site type than in all solvent exposed positions receive a propensity score of greater than 1 , while less frequently occurring types receive a score of less than 1 . The statistical significance of the propensity values was estimated by a bootstrap resampling procedure to compute 95% confidence intervals , implemented in R ( http://R-project . org ) . Propensity values were considered significant ( ) if the corresponding 95% confidence interval did not include the value of 1 [23] . The conservation of each alignment position was quantified using two scores . The first was simply the number of residue types that occurred at the position . The second was a Shannon entropy-like score that captured how non-uniform the distribution of residue type frequencies was at the position . ( 4 ) Alignment positions that contain only one kind of amino acid residue receive a conservation score of 1 , while those with a uniform distribution of residue types receive a score of 0 . The distributions of conservation scores for each kind of alignment position ( bi-functional , ligand-only , protein-only , or all exposed residues ) were compared using the Kolmogorov-Smirnov and Mann-Whitney tests , as implemented in R ( http://R-project . org ) . Each family was assigned one of seven broad functions by SUPERFAMILY [22]: General , Information , Metabolism , Not Annotated , Other , Extracellular processes or Intracellular processes . The function propensities of families with significantly greater or fewer bi-functional positions than expected by chance were computed relative to the frequency of functions in all families . ( 5 ) Functions that occur more frequently in a particular set of families than in all families , receive a score of greater than 1 . The significance of the function propensity values was estimated by a non-parametric bootstrap resampling procedure to compute 95% confidence intervals , implemented in R ( http://R-project . org ) . Propensity values were considered significant ( ) if the corresponding 95% confidence interval did not include the value of 1 .
|
Proteins function through their interactions with other biological molecules , including other proteins . Often times , these interactions underlie cellular processes that go awry in disease . Therefore , modulating these interactions with small molecules is an active area of research for new drugs to treat diseases and new chemical tools to dissect cellular interaction networks . However , targeting protein–protein interactions has proven to be more challenging than the typical drug targets found on individual proteins . Here , we present a computational approach that aims to help in this challenge by identifying regions of protein–protein interfaces that may be amenable to targeting by small molecules . Through a comprehensive analysis of all known protein structures , we identify closely related proteins that in one case bind a protein and in another case bind a small molecule . We find that a significant number of protein–protein interactions occur through surface regions that bind small molecules in related proteins . These “bi-functional” positions , which can bind both proteins and ligands , will serve as an additional piece of structural information that can aid experimentalists in developing small molecules that modulate protein interactions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"computational",
"biology/macromolecular",
"structure",
"analysis"
] |
2010
|
The Overlap of Small Molecule and Protein Binding Sites within Families of Protein Structures
|
Theories of object recognition agree that shape is of primordial importance , but there is no consensus about how shape might be represented , and so far attempts to implement a model of shape perception that would work with realistic stimuli have largely failed . Recent studies suggest that state-of-the-art convolutional ‘deep’ neural networks ( DNNs ) capture important aspects of human object perception . We hypothesized that these successes might be partially related to a human-like representation of object shape . Here we demonstrate that sensitivity for shape features , characteristic to human and primate vision , emerges in DNNs when trained for generic object recognition from natural photographs . We show that these models explain human shape judgments for several benchmark behavioral and neural stimulus sets on which earlier models mostly failed . In particular , although never explicitly trained for such stimuli , DNNs develop acute sensitivity to minute variations in shape and to non-accidental properties that have long been implicated to form the basis for object recognition . Even more strikingly , when tested with a challenging stimulus set in which shape and category membership are dissociated , the most complex model architectures capture human shape sensitivity as well as some aspects of the category structure that emerges from human judgments . As a whole , these results indicate that convolutional neural networks not only learn physically correct representations of object categories but also develop perceptually accurate representational spaces of shapes . An even more complete model of human object representations might be in sight by training deep architectures for multiple tasks , which is so characteristic in human development .
Understanding how the human visual system processes visual information involves building models that would account for human-level performance on a multitude of tasks . For years , despite the best efforts , computational understanding of even the simplest everyday tasks such as object and scene recognition have been limited to toy datasets and poor model performances . For instance , hierarchical architecture HMAX [1] , once known as “the standard model” of vision [2] , worked successfully on a stimulus set of paper clips and could account for some rapid categorization tasks [3] but failed to capture shape and object representations once tested more directly against representations in the visual cortex ( e . g . , [4–6] ) . Recently , however , deep neural networks ( DNNs ) brought a tremendous excitement and hope to multiple fields of research . For the first time , a dramatic increase in performance has been observed on object and scene categorization tasks [7 , 8] , quickly reaching performance levels rivaling humans [9] . More specifically in the context of object recognition , stimulus representations developed by the deep nets have been shown to account for neural recordings in monkey inferior temporal cortex and functional magnetic resonance imaging data throughout the human ventral visual pathway ( e . g . , [6 , 10 , 11] ) , suggesting that some fundamental processes , shared across different hardware , have been captured by deep nets . The stimulus sets on which DNNs have been tested in these previous studies allow the inference that there is a general correspondence between the representations developed within DNNs and important aspects of human object representations at the neural level . However , these stimulus sets were not designed to elucidate specific aspects of human representations . In particular , a long tradition in human psychophysics and primate physiology has pointed towards the processing of shape features as the underlying mechanism behind human object recognition ( e . g . , [12–15] ) . Cognitive as well as computational models of object recognition have mainly focused upon the hierarchical processing of shape ( e . g . , [1 , 16 , 17] ) . There are historical and remaining controversies about the exact nature of these shape representations , such as about the degree of viewpoint invariance and the role of structural information in the higher levels of representation ( e . g . , [18 , 19] ) . Still , all models agree on the central importance of a hierarchical processing of shape . For this reason we hypothesized that the general correspondence between DNNs representations and human object representations might be related to a human-like sensitivity for shape properties in the DNNs . Here we put this hypothesis to the test through a few benchmark stimulus sets , which have highlighted particular aspects of human shape perception in the past . We first demonstrate that convolutional neural networks ( convnets ) , the most common kind of DNN models in image processing , can recognize objects based upon shape also when all other cues are removed , as humans can . Moreover , we show that despite being trained solely for object categorization , higher layers of convnets develop a surprising sensitivity for shape that closely follows human perceptual shape judgments . When we dissociate shape from category membership , then abstract categorical information is available to a limited extent in these networks , suggesting that a full model of shape and category perception might require richer training regimes for convnets .
If convnets are indeed extracting perceptually relevant shape dimensions , they should be able to utilize shape for object recognition . This ability should extend to specific stimulus formats that highlight shape and do not include many other cues , such as silhouettes . The models have been trained for object recognition with natural images , how would they perform when all non-shape cues are removed ? In order to systematically evaluate how convnet recognition performance depends on the amount of available shape and non-shape ( e . g . , color or texture ) information , we employed the colorized version of the Snodgrass and Vanderwart stimulus set of common everyday objects [20 , 21] . This stimulus set consists of 260 line drawings of common objects that are easily recognizable to human observers and has been used extensively in a large number of studies ( Google Scholar citations: over 4000 to [20]; over 500 to [21] ) . In our experiments , we used a subset of this stimulus set ( see Methods ) , consisting of 61 objects ( Fig 1A ) . Three variants of the stimulus set were used: original color images , greyscale images , and silhouettes . First , we asked 30 human observers ( 10 per variant of the stimulus set ) to choose a name of each object , presented for 100 ms , from a list of 657 options , corresponding to the actual of these objects and their synonyms as defined by observers in [20] . Consistent with previous studies [15 , 21] , participants were nearly perfect in naming color objects , slightly worse for grayscale objects , and considerably worse for silhouettes ( Fig 1B , gray bands ) . Moreover , we found that participants were very consistent in their responses ( Fig 1C , gray bands ) . We then presented three convnets with the stimuli and asked them to produce a single best guess of what might be depicted in the image . A correct answer was counted if the label exactly matched the actual label . We found that all deep nets exhibited a robust categorization performance on the original color stimulus set , reaching about 80–90% accuracy ( Fig 1B , with the best model ( GoogLeNet ) reaching human level of performance . Given that the models have not been trained at all on abstract line drawings , we found it an impressive demonstration of convnet feature generalization . As textural cues were gradually removed , convnets still performed reasonably well . In particular , switching to grayscale decreased the performance by about 15% , whereas a further decrease by 30% occurred when inner gradients were removed altogether ( silhouette condition ) . In other words , even when an object is defined solely by its shape , convnets maintain a robust and highly above-chance performance . Notably , a similar pattern of performance was observed when humans were asked to categorize these objects , suggesting that models are responding similarly to humans but are overall less accurate ( irrespective of stimulus variant ) . To investigate the consistency between human and model responses in more detail , we computed a squared Euclidean distance between the average human accuracy and a model accuracy , and normalized it to the range [0 , 1] , such that a consistency of . 5 means that a model responded correctly where a human responded correctly and made a mistake where a human made a mistake about half of the time ( Fig 1C; see Methods for reasoning behind this choice of consistency ) . Overall , the consistency was substantial and nearly reached between-human consistency for color objects for our best model ( GoogLeNet ) . To visualize the amount of consistency , we depicted The best model’s ( GoogLeNet ) performance on silhouettes against human performance ( Fig 1D ) . The performances are well correlated as indicated by the slope of the logistic regression being reliably above zero ( Fig 1D; z-test on GoogLeNet: z = 2 . 549 , p = . 011; CaffeNet: z = 2 . 393 , p = . 017; VGG-19: z = 2 . 323 , p = . 020 ) . Furthermore , we computed consistency between models and found that for each variant of the stimulus set , the models appear to respond similarly and commit similar mistakes ( the between-model consistency is about . 8 for each pairwise comparison ) , indicating that the models learn similar features . Fig 1D also shows that the models sometimes outperformed humans , seemingly in those situations where a model could take an advantage of a limited search space ( e . g . , it is much easier to say there is an iron when you do not know about hats ) . Overall , however , despite the moderate yet successful performance on silhouettes , it is obvious from Fig 1D that there are quite some stimuli on which the models fail but which are recognized perfectly by human observers . Common mistakes could be divided into two groups: ( i ) similar shape ( grasshopper instead of bee ) , and ( ii ) completely wrong answers where the reason behind model’s response is not so obvious ( whistle instead of lion ) . We think that the former scenario further supports the idea that models base their decisions primarily on shape and are not easily distracted by the lack of other features . In either case , the errors might be remedied by model exposure to cartoons and drawings . Moreover , we do not think that these discrepancies might be primarily due to the lack of recurrent processes in these models since we tried to minimize influences of possible recurrent processes during human categorization by presenting stimuli for 100 ms to human observers . It is also possible that better naturalistic training sets in general are necessary where objects would be decoupled from background . For instance , lions always appear in savannahs , so models might be putting too much weight on savannah’s features for detecting a lion , which would be a poor strategy in the case of this stimulus set . Nonetheless , even in the absence of such training , convnets generalize well to such unrealistic stimuli , demonstrating that they genuinely learn some abstract shape representations . In Experiment 2 , we wanted to understand whether convolutional neural networks develop representations that capture the shape dimensions that dominate perception , the so-called “perceived” shape dimensions , rather than the physical ( pixel-based ) form . In most available stimulus sets these two dimensions are naturally correlated because the physical form and the perceived shape are nearly or completely identical . In order to disentangle the relative contributions of each of these dimensions , we needed stimulus sets where a great care was taken to design perceptual dimensions that would differ from physical dimensions . In 1987 , Biederman put forward the Recognition-by-Components ( RBC ) theory [16] that proposed that objects recognition might be based on shape properties known as non-accidental . Under natural viewing conditions , many object’s properties are changing , depending on lighting , clutter , viewpoint and so on . In order to recognize objects robustly , Biederman proposed that the visual system might utilize those properties that remain largely invariant under possible natural variations . In particular , Biederman focused on those properties of object shape that remain unchanged when the three-dimensional shape of an object is projected to the two-dimensional surface on the eye’s retina , such as curved versus straight object axis , parallel versus converging edges , and so on [23] . Importantly , RBC theory predicts that observers should notice a change in a non-accidental property more readily than an equivalent change in a metric property . Consider , for example , geons shown in Fig 4A , top row . Both the non-accidental and the metric variant differ by the same amount from the base geon ( as measured by some linear metric , such a pixelwise or GaborJet difference ) , yet the non-accidental one appears more distinct to us . Over years , Biederman and others consistently found such preference to hold in a large number of studies across species [25–28] , age groups [29–31] , non-urban cultures [32] , and even in the selectivity of inferior temporal neurons in monkeys [24 , 33] . This idea of invariants has also been shown to play an important role in scene categorization [34] and famously penetrated computer vision literature when David Lowe developed his SIFT ( Scale-Invariant Feature Transform ) descriptor that attempted to capture invariant features in an image [35] . Thus , the sensitivity for non-accidental properties presents an important and well-tested line of research where the physical size of differences between shapes is dissociated from the effect of specific shape differences on perception . We tested the sensitivity for non-accidental properties using a previously developed stimulus set of geon triplets where the metric variant is as distinct or , typically , even more distinct from the base than the non-accidental variant as measured in the metric ( physical ) space . Nevertheless , humans and other species report perceiving non-accidental shapes as more dissimilar from the base than the metric ones , presenting us with a perfect test case where , similar to Exp . 2 , physical shape similarity is different from the perceived one . We evaluated model performance on this set of 22 geons ( Fig 4A ) that have been used previously in behavioral [31 , 32 , 36] and neurophysiological studies . A model’s response was counted as accurate if the response to a non-accidental stimulus was more dissimilar from the base than the metric one . We found that all deep but not shallow or HMAX models ( except for HMAX’99 ) showed a higher than chance performance ( Fig 4B ) with performance typically improving gradually throughout the architecture ( Fig 4C; bootstrapped related samples significance test for deep vs . shallow , one-tailed: p < . 001; deep vs . HMAX: p = . 011 ) . Moreover , deeper networks tended to perform slightly better than shallower ones , in certain layers even achieving perfect performance . Overall , there was not any clear pattern in mistakes across convnets , except for a tendency towards mistakes in the main axis curvature , that is , convnets did not seem to treat straight versus curved edges as very distinct . In contrast , humans consistently show a robust sensitivity to changes in the main axis curvature [31 , 36] . Note that humans are also not perfect at detecting NAPs as reported by [36] . Thus , we do not go further into these differences because the RBC theory and most previous behavioral and neural studies only address a general preference for NAP changes , and hence do not provide a systematic framework for interpreting the presence or absence of such preference for specific NAPs . In the first three experiments , we demonstrated convnet sensitivity to shape properties . However , these convnets have been explicitly trained to optimize not for shape but rather category , that is , to provide a correct semantic label . Apparently , categorization is aided by developing sensitivity to shape . But is there anything beyond sensitivity to shape then that convnets develop ? In other words , to what extent do these networks develop semantic representations similar to human categorical representations over and above mere shape information ? Typically , object shape and semantic properties are correlated , such that objects from the same category ( e . g . , fruits ) share some shape properties as well ( all have smooth roundish shape ) that may distinguish them from another category ( e . g . , cars that have more corners ) , making it difficult to investigate the relative contributions of these two dimensions . To overcome these limitations , Bracci and Op de Beeck [37] recently designed a new stimulus set , comprised of 54 photographs of objects , where shape and category dimensions are orthogonal to each other as much as possible ( Fig 5A ) . In particular , objects from six categories have been matched in such a way that any one exemplar from a particular category would have a very similar shape to an exemplar from another category . Thus , the dissociation between shape and category is more prominent and can be meaningfully measured by asking participants to judge similarity between these objects based either on their shape or on their category . By correlating the resulting dissimilarity matrices to human neural data , Bracci and Op de Beeck [37]found that perceived shape and semantic category are represented in parallel in the visual cortex . We employed this stimulus set to explore how categorical information is represented by convnets . As before , participants were asked to judge similarity among stimuli based either on their shape or on their category . Note that even for categorical judgments , participants were asked to rate categorical similarity rather than divide stimulus set into six categories , resulting in idiosyncratic categorical judgments and consistency between humans not reaching ceiling . First , we found that convnets represented shape fairly well , correlating with perceptual human shape judgments between . 3 and . 4 , nearly reaching the human performance limit ( Fig 5C and 5D ) . Unlike before , the effect was not specific to deep models but was also observed in HMAX and even shallow models . This observation is expected because , unlike in previous experiments , in this stimulus set physical form and perceived shape are well correlated . Instead , the purpose of this stimulus set was to investigate to what extent semantic human category judgments are captured by convnets , since here category is dissociated from shape . We found that all deep but not shallow or HMAX models captured at least some semantic structure in our stimuli ( Fig 5D and 5E; bootstrapped related samples significance test for deep vs . shallow and deep vs . HMAX: p < . 001 ) , indicating that representations in convnets contain both shape and category information . Similar to Exp . 1 , comparable correlations were observed even when the models were provided only with silhouettes of the objects ( no texture ) , indicating that such categorical decisions appear to rely mainly on the shape contour and not internal features . The abundance of categorical information in convnet outputs is most strikingly illustrated in Fig 5B where a multidimensional scaling plot depicts overall stimulus similarity . A nearly perfect separation between natural and manmade objects is apparent . Note that less than a half of these objects ( 23 out of 54 ) were known to GoogLeNet , but even completely unfamiliar objects are nonetheless correctly situated . This is quite surprising given that convnets were never trained to find associations between different categories . In other words , there is no explicit reason why a convnet should learn to represent guitars and flutes similarly ( the category of “musical instruments” is not known to the model ) . We speculate that these associations might be learned implicitly , since during training objects of the same superordinate category ( “musical instruments” ) might co-occur in images . Further tests would be necessary to establish the extent of such implicit learning in convnets . Despite its significance , the correlation with categorical judgments was much weaker than with shape , even after we restricted stimuli to the 23 objects in the ImageNet , meaning that the learned representations in convnets are largely based on shape and not category . In other words , categorical information is not as dominant in convnets as in humans , in agreement with [6] where deep nets were shown to account for categorical representations in humans only when categorical training was introduced on top of the outputs of convnets . ( See also Discussion where we talk about the availability of information in models . )
Our results suggest that a human-like sensitivity to shape features is a quite common property shared by different convnets , at least of the type that we tested . However , the three convnets were also very similar , since all of them very trained on the same dataset and used the same training procedure . Which convnet properties are important in developing such shape sensitivity ? One critical piece of information is offer by the comparison to HMAX models . Despite a similar architecture , in most experiments we observed that overall HMAX models failed to capture shape sensitivity to the same extent as convnets . The most obvious difference lies in the depth of the architecture . There are at most four layers in HMAX models but at least eight layers in the simplest of our convnets , CaffeNet . However , HMAX’99 ( that has two layers ) did not seem to perform consistently worse than HMAX-PNAS ( that has four layers ) . Another important difference is the lack of supervision during training . As has been demonstrated before with object categorization [6] , unsupervised training does not seem to be sufficiently robust , at least the way it is implemented in HMAX . Another hint that supervision might be the critical component in learning universal shape dictionaries comes from comparing our results to the outputs obtained via the Hierarchical Modular Optimization ( HMO ) that was recently reported to correspond well to primate neural responses [10] . For Exps . 2a and 4 , where we could obtain the outputs of the HMO layer that corresponds best to monkey neural data , we found largely similar pattern of results , despite differences in depth , training procedure , and training dataset . The only clear similarity between the tested convnets and HMO was supervised learning . Finally , part of convnet power might also be attributed to the fully-connected layers . Both in CaffeNet and VGG-19 , the critical preference for perceived shape emerges at the fully-connected layers . In GoogLeNet , the preference to perceptual dimensions is typically the strongest at the last layer that is also fully-connected , though earlier layers that are not fully-connected also exhibit a robust preference for perceived shape . Other parameters , such as the naturalness of the training dataset or the task that convnet is optimized for , might also contribute to the representations that convnets develop . In short , the tests and the models that we have included in the present paper provide a general answer to our hypotheses about shape representations in convnets , but there are many specific questions about the role of individual variables that remain to be answered . In the literature , at least two theoretical approaches to shape processing have played an important role: image-based theories [19] , which capitalize on processing image features without an explicit encoding of the relation between them , and structure-based theories [18] , which emphasize the role of explicit structural relations in shape processing . Our results do not necessarily provide support for particular theories of shape processing . Of course , in their spirit convnets are closer to image-based theories since there is no explicit shape representation computed . On the other hand , in Exp . 3 we also found that convnets were sensitive to non-accidental properties even without ever being trained to use these properties . While in principle HMAX architectures can also develop sensitivity to non-accidental properties when a temporal association rule is introduced [43] , the fact that such sensitivity automatically emerges in convnets when training for object categorization provides indirect support that non-accidental properties are diagnostic in defining object categories , as proposed by the RBC theory [16] . Of course , a mere sensitivity to non-accidental properties does not imply that convnets must actually utilize the object recognition scheme proposed by the RBC theory [16] . For instance , according to this theory , objects are subdivided into sets of shape primitives , known as geons , and recognized based on which geons compose that particular object , referred to as a “structural description” of the object . Finding an increased sensitivity for non-accidental properties does not necessarily imply that all these other assertions of the RBC theory are correct , and it does not by itself settle the controversy between image-based and structure-based models of object recognition . While we demonstrate an unprecedented match between convnet representations and human shape perception , our experiments only capture a tiny fraction of the rich repertoire of human shape processing . It is clear from Exp . 1 that despite a strong performance , convnets remain about 20% worse than human observers at object recognition on silhouettes . Given that convnets are already very deep and were trained exhaustively , it may be a sign that in order to bridge this gap , convnets need additional layers dedicated to developing more explicit structural representations . Another , more fundamental limitation is their feedforward architecture . Whereas humans are thought to be able to perform many object and scene recognition tasks in a feedforward manner [44–46] , they are certainly not limited to feedforward processing and in many scenarios will benefit from recurrent processing [47] . The role of such recurrent processes has been particularly debated in understanding perceptual organization , where the visual system is actively organizing the incoming information into larger entities [48] . For instance , monkey neurophysiology revealed that figure-ground segmentation benefits both from feedforward and feedback processes [49] , and many models of segmentation utilize recurrent loops ( for an in-depth discussion , see [50] ) . In contrast , despite their superior object categorization abilities , vanilla convnets show rather poor object localization results , with the top-performing model ( GoogLeNet ) in the ImageNet Large Scale Visual Recognition Challenge 2014 scoring 93 . 3% on a single object categorization task , yet localizing that object with only 73 . 6% accuracy [51] . In other words , we showed that convnets sensitivity to shape that reflects human judgments once the object itself can be easily extracted from the image . However , as soon as segmentation and other perceptual organization processes become more complicated , humans but not convnets can benefit from recurrent connections . Thus , recurrent neural networks , which incorporate the feedforward complexity of the tested convnets , might provide an even better fit to human perception than purely feedforward convnets . Finally , we have also argued in [50] that feedforward architectures such as convnets might be lacking critical mechanisms that could contribute to the initial image segmentation . In our view , high performance at object recognition and shape processing tasks should not be taken as evidence that the “convolution-non-linearity-pooling” stack at the heart of convnets is necessarily the right or the full solution of feedforward visual processing yet . Small modifications to this architecture , such as adding feature map correlations [52 , 53] or performing VLAD [54] or Fisher Vector [55] pooling already provides convnets with the ability to segment input images and represent textures and artistic style , all of which might be the part of feedforward computations in human visual cortex . Taken together , we demonstrated that convolutional neural networks trained for multiclass object categorization implicitly learn representations of shape that reflect human shape perception . Moreover , we showed that convnets also develop abstract semantic spaces independent of shape representations that provide a good , albeit weaker , match to human categorical judgments . Overall , our results provide an important demonstration that convnets are not limited to only extracting objective information from the visual inputs ( such as object category ) but can also represent the subjective aspects of visual information in accordance to human judgments . In other words , our work suggests that convnets might be a good candidate model for understanding various perceptual qualities of visual information .
Studies reported here were approved by the Social and Societal Ethic Committee at KU Leuven ( Exp . 1 and 2b ) and the Massachusetts Institute of Technology’s Committee on the Use of Humans as Experimental Subjects ( Exp . 1 ) . Almost all simulations were run with Python using the psychopy-ext package [56] that provides several simple shallow models and bindings to the Caffe library [57] and to several popular computer vision models ( PHOG , PHOW , HMAX-HMIN , and HMAX-PNAS ) , written in MATLAB/C . For online data collection , we used mturkutils , a Python interface package to Amazon Mechanical Turk . For data collection in in Exp . 2b , we used similarity rating interface in MATLAB from [58] . For data analysis , we used several popular free and open source Python packages , including numpy , scipy , scikits-learn , scikits-image [59] , pandas , seaborn , statsmodels , and NLTK [60] . The code and stimulus sets for all of our simulations are available publicly at https://osf . io/jf42a , except in the cases when the stimulus set is already available online ( links to these stimulus sets are provided in the repository and in the text ) or subject to copyright restrictions ( stimulus set for Exp . 4 ) . For a maximal reproducibility , all results reported in this manuscript can be generated with a single command: python run . py report—bootstrap . All stimuli were scaled to 256×256 px size . Convnets further downsampled these images to their own predefined image sizes ( typically around 224×224 px ) . Stimuli pixel intensities were rescaled to the range between 0 and 1 , where 0 corresponds to a black pixel and 1 corresponds to a white pixel , and , for deep models , the mean of the ImageNet training set was subtracted . No further processing was done . We used three groups of models: shallow , HMAX , and deep . Shallow models consist of a single layer of processing , and all features are built manually ( i . e . , there is no training ) . In contrast , HMAX and deep networks have a hierarchical feedforward architecture and have been trained for object categorization . However , HMAX models are not as deep ( up to four layers ) and have not been trained very optimally ( either by manual feature selection or by imprinting stimulus selectivity ) , whereas deep nets acquire their features through training by backpropagation , which operates on all the weights in the network . We computed correlations between a particular layer of a model and behavioral data by first computing a dissimilarity matrix between stimuli and then correlating the resulting dissimilarity matrices . We also used a one minus a Pearson correlation as a distance between stimuli as a metric . Since these dissimilarity matrices are symmetric , we used only the upper triangle to compute correlations . Both correlations were computed using the following formula: ∑i=1n ( xi−x− ) ( yi−y− ) ∑i=1n ( xi−x− ) 2∑i=1n ( yi−y− ) 2 , where x and y correspond to either the outputs of a given model to two different stimuli ( for the dissimilarity matrix computation ) or the values in the upper triangle of these dissimilarity matrices ( for correlational analyses ) . We also conducted the analyses using a normalized Euclidean distance as a metric for producing dissimilarity matrices , but pattern of results remained the same , indicating that the choice of a metric has little effect on our findings . The upper and lower bounds of the ceiling performance ( shown as a gray band in figures ) were estimated using the procedure described in [70] . The upper bound was estimated by computing the average Pearson correlation between each participant’s dissimilarity matrix and the average across all participants after z-scoring data . The lower bound was estimated by computing the average Pearson correlation between each participant’s dissimilarity matrix and the average across the remaining participants after z-scoring data . In Exp . 1 , the consistency between human and model ( or between two models ) was computed as one minus a normalized squared Euclidean distance between the corresponding accuracy vectors x and y: 1−∑i=1n ( xi−yi ) 2n . This metric is a version of the Matching distance that is used for estimating dissimilarity in binary data generalized to the non-logistic case . We chose this metric due to the largely ( but not fully ) logistic nature of our data ( which was not the case in correlational analyses ) . In particular , this consistency measure is high when most of the responses match , whereas a correlation is very low as there is little variance in the data ( i . e . , mostly 1’s ) . The same consistency measure was used for estimating upper and lower bounds of the ceiling performance based on human accuracy instead of the Pearson correlation as described in the correlational analyses . In order to estimate the reliability of effects in our experiments , we used bootstrapping approach ( number of iterations was always 1000 ) . In Fig 1C and Fig 4 , we computed 95% confidence intervals by resampling with replacement model responses ( that is , the vector of 0’s and 1’s that indicates whether model’s response was correct or not ) , and computing the mean accuracy at each iteration . The 95% confidence interval was the reported as the 2 . 5% and 97 . 5% percentiles of the bootstrapped accuracy distribution To estimate the reliability of correlations in the correlational analyses ( all other figures ) , we computed bootstrapped confidence intervals as described in [6] . In particular , dissimilarity matrices were resampled with replacement , and the resulting matrices where correlated for a total of 1000 iterations . Note that diagonal entries in the original dissimilarity matrix were undefined , so these entries were removed from the resampled matrices as well . Again , as done in [70] , only the upper triangle was used for correlations . In Exp . 2b and Exp . 4 , the bootstrapping procedure was carried out in a stratified manner , such that a resampling was done only within each class of stimuli , as done in [6] . In particular , for shape , stimuli were resampled from the same shape class ( e . g . , a new sample for elongated vertical shapes was sampled from the nine available elongated vertical shapes only ) , whereas for semantical category , stimuli were resampled from the same category ( e . g . , a new sample for fruits category was sampled from the six available fruit images only ) . Resampling without stratification yielded qualitatively comparable results . In Exp . 2a and Exp . 3 , no stratification was used because there were too few stimuli ( Exp . 2a ) or no categories ( Exp . 3 ) . Confidence intervals were computed as the 2 . 5% and 97 . 5% percentiles of the bootstrap distribution . To estimate whether shallow , HMAX , and deep models differed , we used a bootstrapped paired-samples significance test ( independent-samples significance test gave largely similar results ) . For each bootstrap resampling , model performance ( correlation with the behavioral or pixelwise dissimilarity matrices ) was averaged across models within a group , and the difference in average performance was computed for each pair of groups ( 3 pairwise comparisons in total ) . Across iterations , this yielded many such differences , which together formed the distribution used for statistical inference ( one for each pair of groups ) . The percentile of scores below zero in each such distribution of differences was reported as the p-value .
|
Shape plays an important role in object recognition . Despite years of research , no models of vision could account for shape understanding as found in human vision of natural images . Given recent successes of deep neural networks ( DNNs ) in object recognition , we hypothesized that DNNs might in fact learn to capture perceptually salient shape dimensions . Using a variety of stimulus sets , we demonstrate here that the output layers of several DNNs develop representations that relate closely to human perceptual shape judgments . Surprisingly , such sensitivity to shape develops in these models even though they were never explicitly trained for shape processing . Moreover , we show that these models also represent categorical object similarity that follows human semantic judgments , albeit to a lesser extent . Taken together , our results bring forward the exciting idea that DNNs capture not only objective dimensions of stimuli , such as their category , but also their subjective , or perceptual , aspects , such as shape and semantic similarity as judged by humans .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"brain",
"social",
"sciences",
"neuroscience",
"learning",
"and",
"memory",
"perception",
"cognitive",
"psychology",
"mathematics",
"human",
"performance",
"statistics",
"(mathematics)",
"cognition",
"memory",
"vision",
"computer",
"and",
"information",
"sciences",
"object",
"recognition",
"behavior",
"visual",
"cortex",
"psychology",
"confidence",
"intervals",
"anatomy",
"biology",
"and",
"life",
"sciences",
"sensory",
"perception",
"physical",
"sciences",
"cognitive",
"science"
] |
2016
|
Deep Neural Networks as a Computational Model for Human Shape Sensitivity
|
Cholera remains an important public health problem in major cities in Bangladesh , especially in slum areas . In response to growing interest among local policymakers to control this disease , this study estimated the impact and cost-effectiveness of preventive cholera vaccination over a ten-year period in a high-risk slum population in Dhaka to inform decisions about the use of oral cholera vaccines as a key tool in reducing cholera risk in such populations . Assuming use of a two-dose killed whole-cell oral cholera vaccine to be produced locally , the number of cholera cases and deaths averted was estimated for three target group options ( 1–4 year olds , 1–14 year olds , and all persons 1+ ) , using cholera incidence data from Dhaka , estimates of vaccination coverage rates from the literature , and a dynamic model of cholera transmission based on data from Matlab , which incorporates herd effects . Local estimates of vaccination costs minus savings in treatment costs , were used to obtain incremental cost-effectiveness ratios for one- and ten-dose vial sizes . Vaccinating 1–14 year olds every three years , combined with annual routine vaccination of children , would be the most cost-effective strategy , reducing incidence in this population by 45% ( assuming 10% annual migration ) , and costing was $823 ( 2015 USD ) for single dose vials and $591 ( 2015 USD ) for ten-dose vials per disability-adjusted life year ( DALY ) averted . Vaccinating all ages one year and above would reduce incidence by >90% , but would be 50% less cost-effective ( $894–1 , 234/DALY averted ) . Limiting vaccination to 1–4 year olds would be the least cost-effective strategy ( preventing only 7% of cases and costing $1 , 276-$1 , 731/DALY averted ) , due to the limited herd effects of vaccinating this small population and the lower vaccine efficacy in this age group . Providing cholera vaccine to slum populations in Dhaka through periodic vaccination campaigns would significantly reduce cholera incidence and inequities , and be especially cost-effective if all 1–14 year olds are targeted .
The Ganges River Delta and Bay of Bengal , including Bangladesh , are considered the birthplace of cholera and the origin of six of the seven cholera pandemics recorded in modern times[1] . While national population-based estimates of cholera incidence are lacking in Bangladesh , the perception among local experts is that cholera is increasingly becoming an urban disease . Indeed , based on long-term systematic laboratory testing of 2% of all diarrheal patients presenting at the icddr , b ( International Centre for Diarrhoeal Disease Research , Bangladesh ) hospital in Dhaka ( locally known as the “cholera hospital” ) , this hospital provides care and treatment to approximately 140 , 000 patients of all ages in each year[2] . Dhaka has also experienced several large cholera outbreaks in the past two decades , especially during widespread floods . During major floods in 2004 , 2007 and 2009 , icddr , b saw an estimated 30 , 000 or more cholera cases annually and V . cholerae overtook rotavirus and ETEC as the main pathogen found among patients with severe diarrhea presenting at the hospital[3] . Several cholera outbreaks have also recently been documented in urban areas in other parts of the country[4 , 5] . Residents of slums and low-income districts are especially vulnerable to cholera infection . A systematic sampling of every third diarrheal patient coming to the icddr , b hospital from the low-income area of Mirpur in Dhaka City found V . cholerae to be the most common pathogen isolated–accounting for 23% of cases , 70% of whom were severely dehydrated [6] . Overcrowded living conditions , inadequate sanitation , and overstressed water systems that are not keeping up with population growth are key contributors to high cholera incidence in urban slums , with tap water supplies often found to be the source . These water supplies , even if initially treated with chlorination , become contaminated during distribution due to illegal connections , leaky pipes and low or negative water pressure–resulting in the mixing of sewerage and drinking water and a dilution in chlorine levels [4 , 5] . Contamination of water at the household level , due to poor hygiene , is also common . The Government of Bangladesh has increasingly expressed interest in controlling cholera since the late 2000s . The Bangladesh delegation to the Executive Board of the World Health Organization ( WHO ) played a key role in the development and subsequent passage of a resolution by the World Health Assembly in 2011 calling for member states and WHO to strengthen efforts to prevent and control cholera through a series of measures , including the use of oral cholera vaccines ( OCVs ) “where appropriate , in conjunction with other recommended prevention and control methods and not as a substitute for such methods” [7] . That same year , the Bangladesh Ministry of Health and Family Welfare ( MOHFW ) played an active role in planning , implementing and monitoring a mass cholera vaccination feasibility study implemented by icddr , b in the Mirpur area of Dhaka , in which more than 123 , 000 persons one year and above received two doses of the bivalent , whole-cell killed oral cholera vaccine , Shanchol ( produced in India ) , either alone or in combination with the promotion of hand washing and point-of-use safe water treatment interventions [8–10] . The vaccine–administered in two doses at least two weeks apart and licensed for use in persons one year and older–was shown in a clinical trial in Kolkata , India to have an overall efficacy of 65% lasting at least five years [11] and has been used through a global vaccine stockpile to preempt or respond to cholera outbreaks in ten countries from 2013 to October 2016 . The Mirpur study found mass cholera vaccination in this impoverished , high-risk area to be feasible–with an overall estimated coverage rate for two doses of 72% , including 67% in adults 18 and older–and generally well accepted by the population [10] . The interest in cholera vaccination among local policymakers has been enhanced by the technology transfer and clinical development of a vaccine identical to Shanchol by a Bangladeshi private sector producer , Incepta Vaccine Ltd . The vaccine , to be marketed as Cholvax , is anticipated to be licensed by the end of 2018 and to cost less than Shanchol ( which has a public sector price of $1 . 85 per dose for single-dose vials ) . Analyses of the impact and cost-effectiveness of introducing a new vaccine are increasingly being conducted by countries and donors–most prominently the Gavi Alliance–to inform decisions about implementing or supporting vaccine introductions and which vaccination or targeting strategies to use . Such analyses using the TRIVAC cost-effectiveness model developed by the Pan American Health Organization ( PAHO ) have reportedly played an important role in decisions by a number of countries in the Latin America and Caribbean region to introduce Haemophilus influenzae type b ( Hib ) , pneumococcal conjugate and rotavirus vaccines , as well as more recently in nine countries in Europe , Africa and the Middle East in deciding whether or not to introduce rotavirus or pneumococcal vaccines [12 , 13] . Impact and cost-effectiveness analyses can especially be important in the case of a vaccine against a disease like cholera , in which the risk of getting the disease varies greatly by location ( due to differing water and sanitation conditions ) and by age group . Such analyses can therefore help policymakers make evidence-based decisions on whether or not to use oral cholera vaccines and which geographic areas and age groups to target in order to have the greatest impact for the lowest possible cost . The purpose of this study was to estimate the impact , cost and cost-effectiveness of preventive cholera vaccination over a ten-year period in a high-risk population of slum dwellers in the city of Dhaka , Bangladesh in order to assist policymakers and global partners in determining whether OCVs should be used in such populations as one of the tools to reduce the cholera risk , and if so , which vaccination strategies would be most effective and efficient . To enhance the relevance and credibility of the results , this study uses local data to model the effectiveness of vaccination on cholera transmission , as well as for several key epidemiological and economic variables .
This protocol has been approved by the Research Review Committee ( RRC ) and Ethical Review Committee ( ERC ) at the icddr , b . All data used for this component was anonymised . Moreover , no active data has been collected for this component of the project . The assumptions and data used for the major parameters for the analyses are shown in Table 1 . Additional parameters for the effectiveness modeling are shown in Table A1 in the technical appendix ( S1 Appendix ) . To identify areas in Dhaka at high risk of cholera , we analyzed data from the laboratory surveillance of 2% of hospitalized diarrhea patients who visited the main icddr , b hospital in Dhaka and 10% of patients who came to the icddr , b treatment center in Mirpur from 2011 to 2015 . The data include place of residence . Average annual cholera incidence rates by the sub-districts of Dhaka ( known as thanas ) were then estimated , using thana-level data from the 2011 census for the denominator . A threshold incidence rate of 1 . 5 cases per 1 , 000 per year was used to select high-incidence thanas . Fourteen of the city’s 43 thanas were found to meet this criteria , with an estimated 2015 population of around 3 . 5 million ( supplemental S1 Table ) . Not all hospitalized cases of cholera in Dhaka are treated at the two icddr , b facilities , and thus this method may underestimate the number of high-incidence thanas , as well as their incidence . However , Mirpur is amongst the areas from where the highest numbers of cholera patients seek treatment at the icddr , b hospitals [6] . To further target those at the highest risk of cholera , the study selected for vaccination slum populations within the 14 high-risk thanas , under the assumption that the vast majority of hospitalized cholera cases come from slum areas . The slum population is assumed to make up around 40% of the total population of these thanas ( or around 1 , 383 , 400 people ) , based on an estimate from the Center for Urban Studies [14] . The study assumes the use of the bivalent whole-cell oral cholera vaccine to be produced locally ( Cholvax ) . The five-year , age-specific vaccine efficacy rates used for two doses of the vaccine are from the Kolkata clinical trial of Shanchol ( 42% for 1–4 year olds , 68% for 5–14 year olds , and 74% for persons 15 years and older ) [11] . The analysis modeled the impact and cost-effectiveness of mass cholera vaccination campaigns for three increasingly large target age groups: 1–4 year olds , 1–14 year olds , and all persons one year and older . These targeting options were selected based on interviews conducted with MOHFW officials and other stakeholders . Although the vaccine has been shown to provide protection for five years , at least for children over five years and adults , the campaigns–to be conducted in two rounds ( one for each dose ) –are proposed to take place every three years over the ten-year period of the analysis . This is to account for population mobility in and out of the slum areas–which has the effect of reducing the population’s vaccination coverage over time–as well as the vaccine’s relatively low efficacy rates in the youngest ( 1–4 year ) age group . All three targeting strategies also include the annual vaccination of new birth cohorts through the routine immunization program to protect them during non-campaign years . The first dose can be provided concurrently with the second dose of measles-containing vaccine scheduled at 15 months of age . We assume a vaccination coverage rate for the two-dose series of 70% for children 1–14 years of age ( for both the routine infant vaccination and campaigns ) and 55% for persons 15 and above . These estimates are based on an average of coverage rates achieved in several OCV campaigns conducted in different countries in recent years , including the Mirpur feasibility study mentioned above ( see S2 Table in supplement ) . A case of cholera is defined in this study as one suffering from acute watery diarrhea requiring a visit to a treatment setting . The estimated average annual incidence rate of cholera requiring treatment in the target slum population is 2 . 3 per 1 , 000 . This rate was derived by applying the proportion of diarrheal cases that were found to be confirmed cholera cases through the systematic testing of patients at the icddr , b Dhaka hospital and Mirpur treatment center to the total number of patients seeking care at the hospital for severe diarrhea[6] . Although the incidence of reported cholera varies by thana , we assume that all high-risk populations in Dhaka have the same high cholera incidence rate and therefore use the same rate for the entire target population . Under the assumption that nearly all cholera cases coming to the icddr , b Dhaka hospital are from slum areas , the denominator was the estimated size of the slum population in Dhaka , based on government population data and the Center for Urban Studies estimate of the percentage of Dhaka residents who live in slums ( 40% ) [14] . Age-specific incidence rates were derived by applying the age distribution of cholera cases found through on-going laboratory-supported cholera surveillance in Matlab from 1997 to 2001 to the overall incidence of 2 . 3/1 , 000 per year . The estimated incidence rates are 7 . 86 per 1 , 000 for children under five years of age , 2 . 65/1 , 000 for 5–14 year olds , and 1 . 38/1 , 000 for persons fifteen and older . Thus , using these estimates , pre-school children are 5 . 7 times more susceptible and school-aged children nearly twice as susceptible of becoming infected with cholera requiring treatment as adults in this population . This increased susceptibility of children might actually reflect increased exposure to cholera due to age-related behavioral differences , but the actual biological mechanism behind this does not affect the model . In the absence of data on cholera case fatality rates , an estimated rate of 1 . 5% was used , based on the opinion of experts at icddr , b . Estimates of the average duration of illness and duration of infection come from the literature [15 , 16] . To estimate the impact of different vaccination targeting strategies on the incidence of cholera in this population over a ten-year period , we used a mathematical model that simulates the dynamics of cholera transmission . This model is based on a previously-published model of cholera transmission in Matlab that used times-series data of cholera incidence and other epidemiological data from Matlab from 1997 to 2001 [17] . Details on the model , including all parameters and differential equations used , are given in the technical appendix ( S1 Appendix ) . The model simulates how a person can be infected by another individual–either symptomatic or asymptomatic–or from the environment ( e . g . , via water ) ( Fig 1 ) . It also simulates the effect of immunity on disease transmission from having been infected ( natural immunity ) or having been vaccinated . In brief , the model places people in one of four compartments: 1 ) susceptible to cholera , 2 ) infected and symptomatic , 3 ) infected but asymptomatic , and 4 ) recovered and immune . The concentration of V . cholerae in the environment ( water ) is tracked in another compartment . Ordinary differential equations are used to model the transition of people between compartments over time , which is affected by such factors as the number of infected persons in the community , the level of V . cholerae in the environment , the time it takes for an infected individual to clear the infection , and the efficacy of either natural or vaccine-induced immunity over time . The model was calibrated to simulate the seasonality of cholera in Matlab over a one-year period . The model assumes four different levels of susceptibility to infection based on age groups ( children under two years of age , 2–4 year olds , 5–14 year olds , and adults 15 and older ) . It also assumes that vaccine efficacy is based on the age at vaccination ( 1–4 , 5–14 and 15+ years old ) , using the age-specific vaccine efficacy estimates from Bhattacharya et al . 2013 described above [11] . In the dynamic model of cholera transmission , the rate of cholera infection is proportional to the number of infected individuals and the amount of Vibrio in the environment . Therefore , vaccination not only reduces the number of cases among vaccinees , but even non-vaccinated individuals are protected indirectly because the averted cases among vaccinees reduce everyone’s risk of infection . The dynamic model was used to estimate the magnitude of this effect . Thus , the indirect ( herd ) protective effects of cholera vaccination are built into this dynamic model . The incorporation of herd effects , along with the simulation of the seasonal pattern of the disease , are meant to provide a more accurate picture of the impact of cholera vaccination on disease incidence over time than would a static outcomes model , in which vaccination does not reduce the incidence of cholera in the unvaccinated population so there are no herd effects . In adjusting the model from Matlab to Dhaka , a migration factor was added to account for the high mobility of slum populations in cities like Dhaka . The model thus replaces a portion of each vaccinated population group with non-vaccinated individuals each year at a constant rate and assumes that these non-vaccinated newcomers have the same level of cholera susceptibility as the baseline ( pre-vaccination ) population . This has the effect of reducing vaccination coverage in the target population over time . Since little data on migration in and out of the slums of Dhaka are available , we modeled three annual migration rates: 0% , 10% and 25% , and used 10% as the base case for the main analyses . However , we assumed that the size of the at-risk population is fixed over the 10-year period of the analysis ( i . e . , as many people leave as enter the targeted areas ) . The output of the model is the number of symptomatic cases of cholera per year , including those not treated , for each age group and each vaccination targeting strategy , varied by migration rates . We then translated the results to adjust for the age structure in Dhaka ( which is somewhat different than that in Matlab ) , based on census data . We assume that most cholera illness is either mild or otherwise unreported , so we computed a "reporting rate" that , when multiplied by the age-adjusted number of symptomatic cases derived from 100 stochastic runs of the model using an unvaccinated population , produced 2 . 3 reported cases per 1 , 000 population per year ( the estimated annual incidence of cholera requiring treatment in the target population , as described above ) . The analysis of cases averted is based on the "reported" number of cases produced by the model . For each vaccination strategy modeled , we obtained the number of cases prevented each year and cumulatively over ten years by subtracting the number of cases predicted once vaccination is implemented from the expected number of cholera cases if no vaccination takes place . The number of deaths averted was calculated by multiplying the number of cases prevented by the assumed case fatality rate of 1 . 5% . Measures of cost-effectiveness were obtained by dividing the net cost of vaccination over the ten-year period of the analysis by the cumulative number of cases , deaths and disability adjusted life years ( DALYs ) prevented as a result of vaccination for each of the targeting strategies . The net vaccination cost is the cost of the vaccination program minus the estimated savings in treatment costs resulting from a reduction in cholera incidence due to vaccination . The resulting incremental cost-effectiveness ratios ( ICERs ) –cost per case averted , cost per death averted , and cost per DALY averted–were calculated for two different vaccine presentations: single-dose vials and ten-dose vials . As is standard , DALYs averted were calculated using DALY weights , a standard discount weight , and life expectancies using methods described in a paper by Fox-Rushby and Hanson [18] . No age weights were used in the analysis . This economic analysis takes a health provider perspective , as opposed to a societal perspective . The vaccination costs are assumed to be paid by the public sector ( government and/or donors ) and only the cost of treating cholera paid by the health care provider are included . Thus , the costs of cholera illness borne by individuals , such as out-of-pocket expenses for medicines , transport and lodging for caregivers and the indirect costs of loss wages of patients or their caregivers from missing work–are not included in the treatment cost estimates . Nor were any private costs related to vaccination , such as the cost of transportation or of missing work to get vaccinated . The cost-of-illness from a societal perspective would include these private costs–resulting in greater treatment savings–and thus our cost-effectiveness measures will be slightly more conservative than if a societal perspective was used . Cost-effectiveness thresholds based on a country’s per capita gross domestic product ( GDP ) have often been used as a measure of the cost-effectiveness of a health intervention , with a cost per DALY averted that is equal to or less than the GDP per capita indicating that the intervention is “very cost-effective” [19] . However , these thresholds have been criticized as too limited as the sole or even a major determinant in decision-making , especially since they do not take into account a country’s specific context , including its ability to afford the intervention [20] . Therefore , in addition in comparing the cost per DALY averted results to per capita GDP , we also examine the affordability of cholera vaccination , in terms of cost per vaccinees and program cost as a percentage of the routine EPI budget . A univariate deterministic sensitivity analysis was conducted to show which variables have the greatest impact on cost-effectiveness . In the Excel spreadsheet that calculated cost-effectiveness , we designated distribution functions for the variables with uncertainty and then ran the Monte Carlo simulations . The gamma distribution function was used to estimate three variables: unit vaccine cost , vaccine delivery cost , and cholera treatment cost , while the beta distribution function was used to estimate two variables that had values that were between 0 and 1: the case fatality rate and cholera incidence . This analysis varies the values of key input parameters–case fatality rate , cholera incidence rates , vaccine price , cost of treatment , and vaccine delivery cost–one by one to estimate how these affect the outcome . Monte Carlo simulations were conducted multiple times ( 10 , 000 ) by drawing random values from the distribution functions for the input parameters using Ersatz software ( version 1 . 3 ) . Two distribution functions are used to model uncertainty: 1 ) beta for incidence and case fatality rates , variables with values between 0 and 1; and 2 ) gamma for vaccine , delivery , and treatment costs . For the gamma distribution , the parameters were a shape parameter α and a mean parameter β . For the beta distribution , the parameters are two positive shape parameters , denoted by α and β , that appear as exponents of the random variable and control the shape of the distribution . The input variable that is the most influential on cost per DALY averted is the one with the longest confidence interval , as will be shown in the tornado graph .
Using a base case of 10% annual migration ( i . e . , the target population is replaced at a rate of 10% a year ) , Fig 2A shows the predicted number of reported cholera cases each year in the target population of around 1 . 4 million people once the vaccination program–consisting of mass vaccination campaigns every three years and annual vaccination of the new birth cohort–is implemented , for each vaccination strategy . 2B depicts the total number of cases for the ten-year period by vaccination strategy and age group . The model was at equilibrium when interventions were simulated , so the simulations with no vaccination also reflect the incidence of cholera before vaccination . The percent reduction in incidence for each vaccination strategy is shown in Fig 3 , while the cumulative number of cases prevented over the ten-year period is shown in Fig 4 . Results showing 95% confidence intervals are presented in S1 Appendix . If no vaccination or other cholera intervention program is enacted , there would continue to be , on average , around 3 , 200 cases of cholera in the study population presenting at health facilities each year , or more than 32 , 000 cases over the ten-year period of the analysis . A strategy of vaccinating only children 1–4 years of age would reduce cholera incidence in the overall targeted population by around 7% − preventing around 2 , 411 cases over ten years or 241 cases per year on average ( Figs 3 and 4 and Table 2 ) . Expanding the target vaccination group to 1–14 year olds would prevent around 14 , 400 cases over ten years , reducing incidence in the overall population by 45% . Vaccinating all ages one and above would prevent more than 29 , 100 cases over this period , reducing the overall cholera burden in the target population by 91% . Viewed from another perspective , for every reported case prevented in the overall population of nearly 1 . 4 million , 123 children 1–4 years of age would need to be vaccinated compared to 80 children 1–14 years old and 102 persons one year and above ( Table 2 ) . The strategy of targeting 1–14 year olds is thus the most efficient . The herd effects of cholera vaccination are clearly shown in these results . The reduction in cases overall and in each age group is modest when vaccination is limited to 1–4 year olds , who make up around 7% of the total study population . However , expanding vaccination to all 1–14 year olds not only reduces the number of cases amongst these children , it also reduces incidence in adults ( who are not vaccinated ) by 40% ( Fig 3 ) . Thus , while 1–14 year olds account for around 30% of the population , vaccinating them would cut cholera incidence in the entire population by 45% . When adults are also vaccinated , more than 90% of cholera cases would be prevented . We tested these results with different levels of annual population migration , which dilutes vaccination coverage of the population over time . When the entire population one year and above is targeted for vaccination and there is no migration , cholera transmission virtually stops and the overall effectiveness of the program is nearly 100% ( Figure S3 in S1 Appendix ) . We also simulated vaccination campaigns every 5 years instead of every 3 years . In these scenarios , we assumed that vaccine protects individuals for 5 years . Vaccinating every 5 years is somewhat less effective than vaccinating every 3 years when 10% annual migration is assumed ( Figure S5A and S5B in S1 Appendix ) , and becomes less effective when the migration rate is high ( Figures S5C and S5D ) . If annual migration is 25% , the effectiveness of this strategy may fall below 90% , while when migration is 10% a year , effectiveness is above 90% . The effectiveness of vaccinating 1–14 year olds on the overall cholera incidence in the population is reduced from nearly 48% at 0% migration to 45% at 10% migration and to 41% at a 25% migration rate . When only 1–4 year olds are vaccinated , the effectiveness is around 7% , regardless of the migration level . The projected costs of vaccination for the ten-year period by vaccination strategy and vial size are shown in Table 3 . Including annual vaccination of infants in all scenarios , the total costs are estimated at $1 . 5 - $1 . 9 million ( depending on the vial size ) for the strategy targeting 1–4 year olds , $4 . 4 - $5 . 9 million for the strategy targeting 1–14 year olds , and $10 . 8 - $14 . 3 million if all persons one year and older are targeted . The annual costs over the ten-year period therefore range from approximately $145 , 600 to $1 . 08 million if ten-dose vials are used and from $193 , 080 to $1 . 4 million if single-dose vials are used . The cost per vaccine recipient , including vaccine wastage , is estimated at $4 . 62 for single-dose vials and $3 . 48 for ten-dose vials . Table 4 shows the net cost of vaccination once the savings in treatment costs resulting from the program are subtracted from the total vaccination program costs . The estimated savings in treatment costs for the ten-year period range from around $126 , 000 , if the program is limited to 1–4 year olds , to $1 . 5 million if all persons one year and above are included . The results of dividing the net vaccination costs by cases , deaths , and DALYs averted for each vaccination strategy and vaccine vial size are shown in Table 5 . The option of vaccinating 1–14 year olds would be by far the most cost-effective , with a cost per case averted of $255 - $356 and a cost per DALY averted of between $591 and $823 . The next most cost-effective strategy would be vaccinating all ages one year and above , with a cost per case averted of $318 - $439 and cost of DALY averted of $894 - $1 , 234–1 . 5 times higher than the 1–14 year old targeting strategy . On the other hand , limiting cholera vaccination to 1–4 year olds would cost $1 , 276 - $1 , 731 per DALY averted–making it 2 . 2 times less cost-effective than the strategy of targeting 1–14 year olds and 1 . 4 times less cost-effective than the option of vaccinating all persons one year and above . Comparing the cost per DALY averted to the GDP per capita as a measure of cost-effectiveness , all vaccination strategies would cost less per DALY averted than the country’s per capita GDP threshold ( $1 , 359 ) [23] , except the option of vaccinating 1–4 year olds only using single-dose vials , indicating that they would be cost-effective using this definition ( Fig 5 ) . However , the 1–4 year old vaccination strategy using ten-dose vials , as well as the strategy of vaccinating all ages one and above using single-dose vials barely fall below the threshold . Univariate deterministic sensitivity analyses were conducted on variables with the greatest uncertainty in their values: case fatality rate ( CFR ) , vaccine price , delivery cost , treatment cost and cholera incidence . The cost per DALY averted estimates vary most with changes in CFR , cholera incidence rates , and cost of treatment , while varying vaccine price and delivery costs has less effect on incremental cost-effectiveness ratios ( Table 6 and Fig 6 ) . However , even when any of the five parameters are varied , the strategy targeting 1–14 year old children for vaccination is the most cost-effective , regardless of the vaccine vial size , followed by vaccination of all persons one year and older . Table 7 shows the mean value and 95% confidence interval for the cost per DALY averted by age group and vial size when the uncertain parameters are varied in the Ersatz Monte Carlo simulations . The results again show that the lowest cost per DALY averted is found when the age group 1–14 is vaccinated using either vial size . Affordability and the budgetary impact of cholera vaccination is another critical factor in assessing the value for money of this intervention . The cost per vaccinee ( vaccine and service delivery ) in this analysis is $3 . 48 if ten-dose vials are used and $4 . 62 if single-dose vials are used–regardless of the age group targeted for vaccination ( Table 3 ) . The annual cost of vaccination in the proposed areas of Dhaka for the most cost-effective strategy targeting 1–14 year olds − $443 , 924 if ten-dose vials are used , and $588 , 715 if single-dose vials are used–represents 0 . 15% to 0 . 2% of the 2018 routine EPI budget of nearly $300 million [24] .
This analysis predicts that a program in which all children 1–14 years of age are targeted in mass cholera vaccination campaigns every three years , coupled with annual vaccination of infants 12–15 months old through the routine immunization program , would reduce cholera incidence by 45% over 10 years in a population of around 1 . 4 million slum dwellers in high-risk areas of Dhaka , Bangladesh , preventing 14 , 400 cases , including in many adults , and cost $4–6 million over ten years . This would be by far the most cost-effective of the three program options , as well as the most efficient , in terms of the number of vaccinations per case averted . Vaccinating all ages one and above would , on the other hand , reduce cholera incidence in this population by more than 90% over 10 years , preventing more than 29 , 000 reported cases . However , this strategy would be considerably less cost-effective than the one targeting 1–14 year olds , and would cost , on average , $1 . 1–1 . 4 million per year or $11–14 million over ten years . The option of vaccinating only young children ( 1–4 year olds ) would have minimum impact on the overall cholera incidence ( 7% reduction ) and despite its lower costs ( $1 . 5 - $1 . 9 million ) , it would be the least cost-effective of the three strategies analyzed . Moreover , if we target 1–14 years age group a total of 14 , 430 cases will be averted whereas 2 , 411 cases will be averted if we target 1–4 years age group . However , these very young children , who are most vulnerable to severe cholera , would be included in the middle option targeting 1–14 year olds . The results for the 1–14 year old and all-ages strategies show the power of herd effects from cholera vaccination , given that the analysis assumes that only a portion of the targeted population would receive the vaccine ( 70% of children and 55% of adults ) and since direct protection from the vaccine is quite modest ( ranging from 42% in children under five to 68% in 5–14 year olds over five years ) . Our results are in general agreement with observations of herd immunity observed in large trials [8 , 25] . A large-scale cluster randomized trial of cholera vaccination was implemented in Dhaka , and effectiveness was modest ( 37% over 2 years ) [8] . The effectiveness could have been reduced by both the high migration rate and contamination between vaccinated and unvaccinated clusters . The large-scale vaccination campaigns represented by our modeling results are likely to result in higher effectiveness than cluster-randomized trials . These findings reinforce those of past cost-effectiveness analyses of cholera vaccination conducted in Bangladesh and elsewhere , which also found that a strategy of vaccinating children was more cost-effective than a strategy that included vaccinating adults–a key factor being the much higher incidence rates of endemic cholera in children than in adults [24 , 26 , 27] . However , the results also show that limiting vaccination to children under five–those typically targeted by national immunization programs and mass vaccination campaigns–would have a minimal impact on cholera incidence and would be considerably less cost-effective than the other two targeting strategies . Key reasons are the lower efficacy of the vaccine in this youngest age group ( 42% ) , and the low level of herd protection from vaccinating such a small proportion ( ≈7% ) of the general population . Although the efficacy of OCV among children under 5 years old has been found to be about half that of older children in endemic settings in several studies [28–30] , the our model could underestimate the impact of vaccinating young children if they contribute more to transmission than older children and adults . A strength of this study is the identification of high-risk areas in Dhaka that would be strong candidates for a targeted cholera vaccination program in Dhaka . The study also improves upon previous cost-effectiveness analyses by using local data on the cost of recently-conducted vaccination campaigns . However , a number of limitations of the analysis need to be pointed out . The mathematical model was calibrated to the dynamics of cholera transmission in the rural area of Matlab . The transmission dynamics and epidemiological patterns of cholera ( e . g . , survival of V . cholerae in the environment , seasonal patterns , age distribution of cases , ratio of asymptomatic to symptomatic infections ) may differ somewhat in urban slums , such as those in Dhaka . Although the population of Matlab and Dhaka differ , we believe that the attack rates of cholera in these highly vulnerable populations subject to seasonal monsoon-driven epidemics are similar . Uncertainty around these parameters , as well as those governing the dynamics of V . cholerae in the environment , is difficult to resolve and can skew projections of outbreak sizes [31] . However , the mathematical model of cholera transmission was used solely to estimate the magnitude of indirect protection from mass vaccination , which should be robust to variation in these parameters as long as the disease attack rates are calibrated . In addition , in the absence of population-based cholera incidence data , an incidence estimate of treated cases ( 2 . 3/1 , 000 ) was extrapolated from data from two hospitals run by icddr , b . While the majority of cholera cases requiring care in Dhaka are reportedly treated at these two facilities , the omission of cases treated elsewhere would have the effect of under-estimating the true cholera incidence rate , making our analyses conservative . On the other hand , the estimated incidence rate is based on the assumption that all cases coming to these hospitals are from the estimated population who live in slums , which may not be the case . There are also uncertainties with the cholera case fatality rate and level of migration in Dhaka . The uncertainty of these key parameters ( CFR , incidence and migration rates ) was addressed in the sensitivity analyses , which in general show that the results , when comparing the different targeting strategies , remained similar when the values of these parameters were varied . Concerning the economic analyses , the average cost of treating hospitalized cases of cholera in Dhaka was taken from a cost study at the main icddr , b hospital , which may have higher costs of treatment ( and a higher standard of care ) than in other health facilities in the city . In addition , the same cost estimate was applied to all cases , regardless of age , while the cost of treating children may differ somewhat from that of treating adults , as found in past studies of the cost of cholera illness [22 , 32] . However , the sensitivity analysis showed that even varying the cost of treatment estimates by a factor of two in either direction did not substantially change the incremental cost-effectiveness ratios . It should also be noted that this analysis focuses on cholera vaccination in a highly-targeted population of around 1 . 4 million people considered at highest risk of cholera , in a city of 8 . 5 million , with more than 18 million in the Greater Dhaka Area [33] . Since there are many slums–and cases of cholera–beyond those in the 14 thanas included in this analysis , controlling or even eliminating cholera in Dhaka would likely require expanding the target population to other areas in ( and possibly surrounding ) the city . This would increase the costs of the program and could also reduce its cost-effectiveness , if the cholera incidence rates in other areas are lower than the rate used in this analysis . The producer of Cholvax has recently indicated that the public sector price for single-dose vials could be reduced to as low as $1 . 10 ( vs . $1 . 40 assumed in the base analysis ) . This lower price would significantly reduce the costs and thus increase the cost-effectiveness of cholera vaccination under the modeled scenarios . Cost-effectiveness is not the only factor that policymakers and donors need to consider when making decisions about whether or not to implement cholera vaccination in high-risk areas and which age groups to target . Another key consideration is affordability , as the financial realities in Bangladesh make it unlikely that the country can readily implement all potentially cost-effective interventions . The most cost-effective strategy–that of vaccinating 1–14 year olds in this population of around 1 . 4 million people–would cost , on average , ≈$444 , 000 - $589 , 000 per year , while the most expansive strategy of targeting all ages one and above would cost around 60% more ( $1 . 1–1 . 4 million ) . The 1–14 year old option would thus add 0 . 15%– 0 . 2% to the annual national immunization program budget ( of around $300 million ) , while the all-ages strategy would add 0 . 4%-0 . 6% . Assuming vaccine protects for 5 years for all age groups , we found that spacing the campaigns to every 5 years led to more cases 4 or 5 years after campaigns but only decreased average effectiveness over 10 years only a modest amount . However , the 5-year campaigns were less effective when migration was higher . If protection from vaccination wanes more rapidly among young children [30] , then less frequent vaccination campaigns might not adequately protect them . However , expanding the vaccination program to other areas of Dhaka or to other cities in Bangladesh would increase these costs . In terms of the estimated cost per vaccinee − $4 . 62 for single-dose vials and 3 . 48 for ten-dose vials–this compares to a cost estimate for rotavirus vaccination ( using Rotarix ) of $5 . 46 - $5 . 98 per child vaccinated from an analysis conducted by PATH ( Clint Pecenka , personal communications ) . If protection from vaccination wanes more rapidly among young children , then less frequent vaccination campaigns might not be effective . One way to reduce the costs of cholera vaccination , and thus increase its affordability and cost-effectiveness , would be to decrease the frequency of the vaccination campaigns from every three years to every five years , given that the vaccine has been shown to provide protection for at least five years , at least among persons five years and older . However , in a highly mobile area such as the slums of Dhaka , population migration would erode coverage and decrease the effectiveness of vaccination over time . If one assumes , for example , an annual migration rate of 10% , 50% of the population will have been replaced with unvaccinated people within five years , greatly reducing the level of protection in this population and potentially leading to outbreaks three or four years after a campaign . Conducting vaccination campaigns every five years may be an appropriate strategy in a less mobile population . Government policymakers in Bangladesh have expressed interest in providing cholera vaccination in combination with other interventions to reduce cholera and water-borne diseases , as recommended by WHO [34] . A comprehensive cholera control program that combines vaccination with improvements in water distribution and water quality–such as by increasing the number of legal water connections and placing water pipes far from sewer systems—should , in fact , create synergies that result in a more rapid reduction in disease than any intervention on its own [35] . Cholera vaccination could also be a means of accelerating control of the disease before adequate water and sanitation improvements can reach the most vulnerable populations . In addition , cholera vaccination could be provided during periodic intensive routine immunization ( PIRI ) activities or other vaccination campaigns ( e . g . , measles-rubella ) , which would further reduce its costs . Another key factor to consider in deciding whether or not to introduce a new vaccine is fairness and equity [20] . Because cholera predominantly strikes the most impoverished and marginalized populations , who are also those with the least access to quality health care services , cholera vaccination , especially using a strategy targeting high-risk areas , would reduce these inequities . While cholera incidence rate estimates are not available for other urban areas of Bangladesh , it is likely that in cities where the disease is known to be endemic or where outbreaks have occurred in the recent past and where slum conditions are similar to those in Dhaka , the risk of cholera will be similar to that found in the slums of Dhaka . Thus , it is reasonable to assume that the results of this analysis can be generalized to slum populations in other cholera-affected urban areas of the country . The findings from icddr , b’s prospective cholera surveillance currently underway in twenty-two mainly urban sites throughout the country will help increase our understanding of the cholera burden in other parts of the country and thus determine the relevance of the findings of our analysis to other urban areas . This strategy of targeting urban slums for special disease control efforts would also align with a new national government priority of improving health in slums in Bangladesh .
|
While oral cholera vaccines are increasingly being used in the past few years , mainly to curtail or preempt cholera outbreaks , they have yet to be used on a large scale to control endemic cholera in a high-burden country like Bangladesh . This study examines the potential impact on disease and value of vaccinating slum dwellers in Dhaka ( and by extension in other cities ) , who are among those at highest-risk of getting the disease . This analysis suggests that , despite the modest efficacy and limited duration of protection of existing vaccines , mass cholera vaccination can have a significant impact on reducing cholera incidence in the entire population–including those not vaccinated–as a result of herd effects–and can be a cost-effective means of controlling the disease , especially until more long-term measures , such as improved water and sanitation infrastructure , are put in place . These results should assist policymakers and potential donors in determining whether and how to use these vaccines in Bangladesh to control the disease amongst its most vulnerable populations .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"cost-effectiveness",
"analysis",
"economic",
"analysis",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"vaccines",
"bacterial",
"diseases",
"age",
"groups",
"animal",
"behavior",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"zoology",
"bangladesh",
"infectious",
"diseases",
"cholera",
"cholera",
"vaccines",
"behavior",
"economics",
"people",
"and",
"places",
"psychology",
"animal",
"migration",
"asia",
"population",
"groupings",
"biology",
"and",
"life",
"sciences"
] |
2018
|
The impact and cost-effectiveness of controlling cholera through the use of oral cholera vaccines in urban Bangladesh: A disease modeling and economic analysis
|
Chronic infection with Trypanosoma cruzi leads to a constant stimulation of the host immune system . Monocytes , which are recruited in response to inflammatory signals , are divided into classical CD14hiCD16— , non-classical CD14loCD16+ and intermediate CD14hiCD16+ subsets . In this study , we evaluated the frequencies of monocyte subsets in the different clinical stages of chronic Chagas disease in comparison with the monocyte profile of seronegative heart failure subjects and seronegative healthy controls . The effect of the anti-parasite drug therapy benznidazole on monocyte subsets was also explored . The frequencies of the different monocyte subsets and their phenotypes were measured by flow cytometry . Trypanosoma cruzi-specific antibodies were quantified by conventional serological tests . T . cruzi-infected subjects with mild or no signs of cardiac disease and patients suffering from dilated cardiomyopathy unrelated to T . cruzi infection showed increased levels of non-classical CD14loCD16+ monocytes compared with healthy controls . In contrast , the monocyte profile in T . cruzi-infected subjects with severe cardiomyopathy was skewed towards the classical and intermediate subsets . After benznidazole treatment , non-classical monocytes CD14loCD16+ decreased while classical monocytes CD14hiCD16—increased . The different clinical stages of chronic Chagas disease display distinct monocyte profiles that are restored after anti-parasite drug therapy . T . cruzi-infected subjects with severe cardiac disease displayed a profile of monocytes subsets suggestive of a more pronounced inflammatory environment compared with subjects suffering from heart failure not related to T . cruzi infection , supporting that parasite persistence might also alter cell components of the innate immune system .
Chagas disease , caused by infection with the intracellular protozoan parasite Trypanosoma cruzi , affects 6–7 million people and represents the most frequent cause of infectious cardiomyopathy in the world [1 , 2] . Three factors are likely associated with the development of the more severe clinical forms of the disease: parasite burden , the capacity of the host immune response to control parasites in specific tissues , and the effectiveness of the host immune response to control tissue damage . In response to inflammatory signals , circulating monocytes leave the bloodstream and migrate into tissues , where following conditioning by local growth factors , pro-inflammatory cytokines and microbial products , they differentiate into macrophages or dendritic cells . Although the recruitment of monocytes is essential for the effective control and clearance of microorganisms , they can also be highly damaging to neighboring tissues [3] . Human monocytes are divided into subsets on the basis of surface CD14 and CD16 expression [4] . CD14hiCD16−monocytes “classical Mo” , which are also referred to as classical monocytes , are the most prevalent monocyte subset in human blood , and they show a high expression of the chemokine receptor CCR2 . Classical monocytes can migrate to sites of injury and infection , where they differentiate into inflammatory macrophages [5] . The CD16+ monocyte population comprises two subsets: the non-classical CD14loCD16+ “non-classical Mo” and the intermediate CD14hiCD16+ monocytes “intermediate Mo” [4 , 6] . Both subsets exhibit low and mild CCR2 expression [7] . Whereas non-classical monocytes are involved in the process of patrolling with potent anti-inflammatory function and wound healing , intermediate monocytes share some phenotypic and functional features of both classical and non-classical monocytes and mainly exert a pro-inflammatory role [7] . The two CD16+ subsets are shown to expand in many inflammatory conditions ( e . g . , cancer , sepsis and stroke ) and infections such as HIV [8–11] and tuberculosis [12] . In the chronic phase of Chagas disease , T cell responses become exhausted over time , presumably due to the constant stimulation of the host immune system in this decades-long infection [13–15] . This constant stimulation of the host immune system is also evident by the expansion of CD14+CD16+HLA-DR++ monocytes [16] , which shows that adaptive and innate immune responses can be disrupted in the chronic phase of the infection . Chronic Chagas heart disease presents morphological particularities that could account for a worsened clinical course compared with dilated cardiomyopathy not related to T . cruzi infection [17] . Herein , we sought to evaluate the frequencies of classical , intermediate and non-classical monocytes in the different clinical stages of chronic Chagas disease compared with the monocyte profile in seronegative dilated cardiomyopathy patients ( DCM ) and seronegative healthy controls . The effect of the anti-parasite drug therapy benznidazole on monocyte subsets was also explored in chronically infected subjects .
The protocol was approved by the institutional review boards of Hospital Interzonal General de Agudos Eva Perón , Buenos Aires , Argentina . Signed informed consent was obtained from all individuals before inclusion in the study . T . cruzi-infected adult volunteers were recruited at the Chagas Disease Unit of Hospital Interzonal General de Agudos Eva Perón , Buenos Aires , Argentina . T . cruzi infection was determined by indirect immunofluorescence assay , hemagglutination assay , and enzyme-linked immunosorbent assay ( ELISA ) in compliance with domestic and international criteria [1] . The ELISA was carried out with a 1/200 dilution of the samples incubated in microplates precoated with T . cruzi epimastigote antigens . The binding of specific antibodies was detected with a horseradish peroxidase-labeled anti-human IgG antibody ( Sigma ) . After addition of the substrate o-phenylenediamine ( Sigma ) , the optical density at 490 nm ( OD490 ) was quantified in an ELISA microplate reader ( Model 550; Bio-Rad , Tokyo , Japan ) [18] . The chronically infected seropositive subjects were clinically evaluated and stratified according to a modified version of the Kuschnir grading system [19 , 20] . The individuals in group 0 had normal electrocardiographic ( ECG ) , normal chest radiographic , and normal echocardiographic findings; the subjects in group 1 had normal chest radiographic and echocardiographic findings but abnormal ECG findings; the subjects in group 2 had ECG abnormalities and heart enlargement; and the subjects in group 3 had ECG abnormalities , heart enlargement , and clinical or radiological evidence of heart failure . A group of seronegative subjects suffering from DCM with systolic heart failure were recruited for comparison of the monocyte subset phenotypes among patients with heart failure due to different disease etiologies . The inclusion criteria for patients with heart failure were class I/II/III ( New York Heart Association classification ) , with an ejection fraction of <40% by echocardiography . The etiology for heart failure was hypertension in three patients , post-chemotherapy with doxorubicin in one patient who had no cancer at the time of study inclusion , alcoholism in one patient , and idiopathic DCM in three patients . Seronegative ( uninfected ) healthy controls were also included . Subjects with acute coronary syndrome , cancer , HIV , syphilis , diabetes , arthritis , or serious allergies at the time of study inclusion were excluded . At the time of the study , all of the participants were living in Buenos Aires , where T . cruzi infection is not endemic . After inclusion in the study , nine T . cruzi-infected subjects in the G0 group and three subjects in the G1 clinical group were treated with 5 mg·kg per day of benznidazole for 30 days [21 , 22] . Clinical , serological , and immunological analyses were performed prior to and at different time points after the treatment . Data on the number , sex , and age of the enlisted subjects are summarized in Table 1 . Whole blood was drawn by venipuncture into heparinized tubes ( Vacutainer; BD Biosciences ) . PBMCs were isolated by density gradient centrifugation on Ficoll-Hypaque ( Amersham ) and diluted in RPMI media containing 10% newborn bovine serum , 100 units/ml penicillin , 0 . 1 mg/ml Streptomycin , 2 mM L-glutamine and 10 mM HEPES buffer . The viability of the cells was checked by trypan blue staining with a viability range of 80–90% . A blood aliquot was allowed to coagulate at room temperature and centrifuged at 1000 g for 15 min for serum separation . Immediately after collection , 1 × 106 PBMCs were stained with different combinations of FITC-labeled anti-CD14 , PE-labeled anti-CD16 , APC-labeled anti-CD45RA and AF647-labeled anti-CCR2 ( all from BD Pharmingen ) at 4°C for 30 min . The cells were then fixed with 2% paraformaldehyde and stored at 4°C until acquisition . The cells were acquired with a BD FACS Calibur flow cytometer ( BD Biosciences ) and analyzed with FlowJo software v9 . 6 ( Tree Star ) . Monocyte subsets were first selected on the basis of forward-scattered ( FSC ) vs . side-scattered ( SSC ) lights , and CD14+ cells were subsequently gated . From this population , CD14 vs . CD16 dot plots were drawn to establish the different CD14+ monocyte subsets ( S1A , S1B and S1D Fig ) . For monocyte phenotyping , histograms for the expression of CD45RA and CCR2 were plotted for each monocyte subset ( S1E Fig ) . Unstained and fluorescence minus one ( FMO ) samples were used as gating controls ( S1C and S1E Fig ) . To demonstrate that the majority of the cells selected by the monocyte gating were truly monocytes , additional analyses were performed as follows . The PerCP-labeled anti-HLA-DR antibody ( Biolegend ) was added to the combination of CD14 and CD16 antibodies mentioned above , and CD14 vs . CD16 dot plots were drawn after the selection of HLA-DR+ cells from the CD14+-gated population [23 , 24] ( S1A , S1B , S1E and S1F Fig ) . Staining with the APC-labeled anti-CD19 , APC-Cy7-labeled anti-CD3 , PE-labeled anti-CD14 , and FITC-labeled anti-CD16 antibodies and with FV510 was performed to ascertain the contribution of any possible contaminating cells , including B , T , to the proportion of the different monocyte subsets . These additional assays were carried out on a FACS Aria II flow cytometer ( BD Biosciences; S1 Fig ) . The demographic and clinical characteristics of T . cruzi-infected subjects included in this study were summarized using the range and median . The normality of data was evaluated by the Shapiro-Wilk test . Differences among groups were evaluated by ANOVA followed by a Bonferroni/Dunn test for multiple comparisons or the Kruskal-Wallis test followed by post-tests , as appropriate . To evaluate the changes in monocyte subsets over time post-treatment compared with the baseline , a linear mixed model with compound symmetry and time as a fixed effect was used to maximize the utilizable data , as some subjects had missing data . The correlation between the frequencies of monocyte subsets and the post-treatment/pretreatment ratio of T . cruzi-specific antibodies measured by ELISA assay was determined by Spearman’s test . The networks comprising the different monocyte subsets were created after performing a correlation analysis by Spearman’s correlation test . Differences were considered statistically significant at P <0 . 05 .
On the basis of the CD14 and CD16 expression levels , CD14+ monocytes were subdivided into classical ( CD14hiCD16− , “classical Mo” ) , intermediate ( CD14hiCD16+ , “intermediate Mo” ) , and non-classical ( CD14loCD16+ , “non-classical Mo” ) subsets ( Fig 1A and S1 Fig ) and were quantified in untreated T . cruzi-infected subjects , seronegative ( uninfected ) patients with DCM , and in seronegative healthy controls . The frequencies of the different monocyte subsets did not change significantly after preselection of HLA-DR+ cells from the CD14+-gated population ( S1A , S1B , S1E and S1F Fig ) . The preselection of HLA-DR+ cells allowed for the exclusion of HLA-DR–negative NK cells [23 , 24] . We confirmed that the contribution of CD3+ and CD19+ cells to the frequencies of the different monocyte subsets selected from the total CD14+-gated population was very low either in T . cruzi-infected or uninfected subjects . As presented in S2 Fig , 1 . 58% of all CD14+ monocytes in an uninfected subject showed positive staining for CD19; this figure represents the final frequency of 1 . 3% , 0 . 12% , and 0 . 094% of classical , intermediate , and non-classical monocytes , respectively . Likewise , 2 . 54% of all CD14+ monocytes stained for CD3; this percentage represents the final frequency of 1 . 84% , 0 . 1% , and 0 . 072% of classical , intermediate , and non-classical monocytes , respectively . The frequency of classical monocytes was higher in T . cruzi-infected subjects with severe cardiomyopathy ( i . e . , the G2 and G3 clinical groups ) than in patients with no signs of cardiac disease ( i . e . , the G0 clinical group ) . Although not statistically significant , patients in groups G2 and G3 had higher frequencies of the CD14hiCD16– “classical Mo” monocyte subset than those of the uninfected healthy controls and G1 and DCM patients ( Fig 1B ) . In contrast , T . cruzi-infected subjects with severe cardiomyopathy had similar frequencies of non-classical CD14loCD16+ monocytes to the uninfected healthy control levels , whereas T . cruzi-infected subjects in the G0 clinical group and patients suffering from DCM unrelated to T . cruzi infection ( DCM ) showed significantly higher levels of non-classical CD14loCD16+ monocytes than the uninfected healthy controls ( Fig 1B ) . A slight increase in non-classical CD14loCD16+ monocytes in G1 patients compared with that in the uninfected healthy controls was also observed . Increased frequencies of intermediate CD14hiCD16+ monocytes were found in chronically infected subjects with more severe stages of the disease and DCM compared with the levels in uninfected healthy controls . The CD14hiCD16+ “intermediate Mo” frequencies were also slightly ( but not significantly ) increased in patients with less severe forms of Chagas disease ( Fig 1B ) . Although patients with chronic Chagas disease with heart failure were older than those in the G0 and the uninfected healthy control groups , we did not find any correlation between the age of the subjects and the frequency of the different monocyte subsets in our study cohort ( classical Mo r = 0 . 219 , P = 0 . 518; intermediate Mo r = –0 . 002 , P = 0 . 996; non-classical Mo r = 0 . 033 , P = 0 . 923; ) . A distinct network profile of monocyte subsets was observed in T . cruzi-infected subjects that varied according to disease severity . The G0 group , with no signs of cardiac disease , had a moderate inverse correlation between classical CD14hiCD16– and non-classical CD14loCD16+ monocyte subsets ( Fig 1C ) . This correlation was not observed in the uninfected healthy controls ( r = -0 . 329 , P = 0 . 198 ) or seronegative subjects with DCM ( r = -0 . 359 , P = 0 . 389 ) . The inverse correlation between classical and non-classical monocyte subsets was increased in the G1 patient group and was sustained in patients with Chagas disease with more severe cardiomyopathy ( Fig 1C ) . The expression of CD45RA in the different monocyte subsets concurred with the expression data reported in other studies [25–27] , and did not vary between patients with chronic Chagas disease regardless the clinical status and uninfected healthy controls ( Fig 2A and S1 Fig ) . In contrast , chronically infected subjects with no signs of cardiac dysfunction had CD14hiCD16– and CD14hi CD16+ monocyte subsets with higher CCR2 expression than those found in patients with severe cardiac disease and in the uninfected healthy controls , respectively ( Fig 2B and S1 Fig ) . To address the relationship between the different monocyte subsets and parasite persistence , classical , non-classical and intermediate monocytes were measured in chronic Chagas disease patients prior to and following treatment with benznidazole . A sharp decrease in non-classical monocytes CD14loCD16+ along with an increase in classical monocytes CD14hiCD16- was observed six months after drug therapy ( Fig 3A ) . Of note , the decrease in non-classical monocytes post-treatment was restricted to those patients who had baseline levels above the median values ( i . e . , non-classical Mo 1 patient group ) , whereas classical monocytes were increased post-treatment in patients who had baseline CD14hiCD16- “classical Mo” frequencies under the median values ( i . e . , classical Mo 2 patient group ) ( Table 2 ) . Likewise , when patients were classified by those who had baseline frequencies of intermediate CD14hiCD16+ monocytes above ( i . e . , intermediate Mo 1 patient group ) or under ( i . e . , intermediate Mo 2 patient group ) the median values , the frequencies of CD14hiCD16+ “intermediate Mo” decreased in the former group , while the other group presented no changes in this monocyte subset after benznidazole therapy ( Table 2 ) . Although the changes were more pronounced at six months post-treatment , the frequencies of classical and non-classical monocytes were in the range of the uninfected healthy controls ( i . e . , classical monocytes , range = 48 . 4%-90 . 7% and non classical monocytes , range = 1 . 56%-12 . 9% ) by 12–24 months following drug therapy . Treatment with benznidazole changed the network profile , inducing a positive correlation between classical and intermediate monocyte subsets and a strong negative correlation between intermediate and non-classical monocyte subsets within 12–24 months of follow-up post-treatment ( Fig 3B ) . We then evaluated whether the treatment efficacy , determined by the presence of significant decreases in T . cruzi-specific antibodies [20] , was associated with a particular profile of monocyte subsets prior to drug therapy . An inverse correlation was observed between the frequencies of classical monocytes prior to treatment and the rate of decreases in T . cruzi-specific antibodies post-treatment ( i . e . , a lower ratio post-treatment vs . pre-treatment in subjects with high baseline frequencies of classical monocytes ) ( Fig 3C , S3 Fig ) . In contrast , the frequencies of non-classical monocytes prior to treatment were positively correlated with the decrease in parasite-specific antibodies post-treatment ( Fig 3C ) ( i . e . , a lower ratio post-treatment vs . pre-treatment in subjects with low baseline frequencies of non-classical monocytes ) .
Although the cause of morbidity in chronic T . cruzi infection has been the source of much debate and controversy , most of the available data support the conclusion that Chagas disease is the result of the failure of the immune system to completely clear this persistent infection and the resulting effects of decades of immune assault [28 , 29] . In the present study , we found that the monocyte profile in chronically T . cruzi-infected subjects varies according to disease severity and changes after anti-T . cruzi treatment with benznidazole . The monocyte profile in less severe forms of cardiac disease is enriched in non-classical monocytes ( CD14loCD16+ ) , whereas the monocyte profile in Chagas disease patients with severe cardiomyopathy is skewed toward classical and intermediate monocytes . These findings suggest that patients without signs of cardiac dysfunction or mild cardiac disease have a more balanced monocyte profile with both proinflammatory and anti-inflammatory tissue repair capacity [5] . Recent studies revealed that classical monocytes exit from bone marrow into the blood stream , where they give rise to intermediate monocytes , which subsequently differentiate into non-classical monocytes [11 , 30] . Other authors have reported that non-classical monocytes may also arise independently from myeloid progenitors in the bone marrow [31] . The inverse correlation between classical and non-classical monocytes in chronically T . cruzi-infected subjects suggests that non-classical monocytes may derive from classical monocytes . Nonetheless , this inverse association may also be due to more active recruitment of classical than non-classical monocytes in T . cruzi-infected tissues in the G0 and G1 groups . In line with these findings , the enhanced expression of CCR2 in classical and intermediate monocytes of T . cruzi-infected patients without signs of cardiac disease may support more active recruitment of these monocyte subsets . Upon activation , classical monocytes produce inflammatory cytokines , may exert phagocytic and myeloperoxidase activities , and release heightened levels of superoxide [32]; these actions altogether can help to keep the parasite under control . Although these responses maintained over time may also result in tissue damage , the recruitment of non-classical monocytes may counteract these harmful effects . In contrast , a more inflammatory environment and tissue damage observed in patients with severe stages of chronic Chagas disease may be responsible for more active chemoattraction and recruitment of non-classical monocytes to sites of inflammation to perform their anti-inflammatory and tissue repair functions [5] , accounting for the stronger inverse correlation between classical and non-classical monocytes in these patients . The extensive fibrosis observed in Chagas disease patients with heart failure might be a consequence of exacerbated remodeling [33] mediated by the tissue repair function of non-classical monocytes . Of note , no correlation between classical and non-classical monocytes was observed in DCM , in agreement with the low-grade inflammation associated with heart failure of non-infectious origin [34] . Several studies suggest that T . cruzi-infected subjects with an indeterminate form of the chronic disease have an overall modulatory cytokine profile of monocytes , with the production of IL-10 as a counterbalance cytokine [16 , 35–40] . In line with our findings , a recent study revealed that chronically T . cruzi-infected subjects with cardiac dysfunction have an increased frequency of intermediate monocytes [41] . Nevertheless , the increased counts of both non-classical monocytes at less severe clinical stages in our study are at odds with the results reported by other authors [40 , 41] . Treatment with benznidazole induced reductions in both non-classical and intermediate monocyte subsets along with an increase in classical-monocyte numbers . This finding is probably due to a reduction in parasite load and the subsequent decrease in inflammation and recruitment of non-classical and intermediate monocytes , thus enabling replenishment of classical monocyte subsets in the circulation . The inverse correlation between non-classical and intermediate monocytes after benznidazole therapy suggests that the signals that induce the rise in the number of non-classical monocytes in the G0 and G1 clinical groups , which are the target populations for trypanocidal treatment , may disappear after this therapy . It has been shown that monocytes can be reprogrammed and switch from a wound-healing to a pro-inflammatory state in response to changes in inflammatory stimuli [42 , 43] . Accordingly , we recently observed that the levels of MCP-1 , one of the main chemokines that regulate migration and infiltration of monocytes and/or macrophages , and of IP-10 , which acts as a modulator of angiogenesis and wound healing , decrease in T . cruzi-infected children after treatment with benznidazole ( Albareda MC , personal communication ) . In contrast to our findings , Sathler-Avelar et al . demonstrated that a pediatric population of T . cruzi-infected subjects treated with benznidazole has higher percentages of non-classical CD14+CD16+ cells as compared with an untreated group [44] . These two studies have several distinct features that might explain these apparent discrepancies , including the difference in the classification of monocytes and in the length of infection . Benznidazole treatment in the indeterminate phase of T . cruzi infection has been reported to downregulate monocyte phagocytic capacity [45] , further supporting the overall low immune activation after benznidazole therapy . An important observation of our work is that the decline in T . cruzi-specific antibody levels after benznidazole therapy , suggestive of successful treatment , was associated with lower baseline levels of non-classical monocytes and higher baseline levels of classical monocytes . A proper balance among the different monocyte subsets may be critical for preventing persistent inflammation and for achieving controlled repair , which might also be an important factor for treatment efficacy . One limitation of this study is that the functional status of the different monocyte subsets was not assessed prior to and after therapy . Additionally , the gating strategy employed did not allow us to exclude the low frequencies of contaminating B and T cells in the different CD14+ monocyte subsets . In summary , T . cruzi-infected subjects with severe cardiac disease have a profile of monocyte subpopulations that is suggestive of a more pronounced inflammatory environment as compared with the subjects without signs of cardiac dysfunction and those with heart failure not related to T . cruzi infection . These findings further indicate that parasite persistence may also alter cell components of the innate immune system .
|
Monocytes are key players during infection , and they leave the bloodstream and migrate into tissues in response to inflammatory signals . Although the recruitment of monocytes is essential for the effective control and clearance of microorganisms , they can also be highly damaging to neighboring tissues . Based on the expression of CD14 and CD16 , monocytes are classified into classical , non-classical and intermediate subsets , all of which exert different functions . Because chronic T . cruzi infection induces a constant activation of the host immune system , inflammatory signals are exacerbated , possibly leading to alterations in the frequencies of monocyte subsets . In this study , we evaluated the monocyte profile in Trypanosoma cruzi-infected subjects with different degrees of cardiac dysfunction and explored whether this profile was similar between seropositive and seronegative subjects with heart failure . We found that the different clinical stages of chronic Chagas disease displayed distinct monocyte profiles , which are susceptible to being restored by modulating the parasite load with anti-parasite drug therapy . T . cruzi-infected subjects with severe cardiac disease displayed a profile of monocytes subsets suggestive of a more pronounced inflammatory environment compared with subjects suffering from heart failure not related to T . cruzi infection , supporting that parasite persistence might be a detrimental factor in the evolution of the cardiac disease induced by T . cruzi .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"inflammatory",
"diseases",
"cardiomyopathies",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"enzyme-linked",
"immunoassays",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoans",
"neglected",
"tropical",
"diseases",
"immunologic",
"techniques",
"research",
"and",
"analysis",
"methods",
"cardiology",
"white",
"blood",
"cells",
"dilated",
"cardiomyopathy",
"animal",
"cells",
"immunoassays",
"protozoan",
"infections",
"heart",
"failure",
"trypanosoma",
"cruzi",
"trypanosoma",
"chagas",
"disease",
"eukaryota",
"cell",
"biology",
"monocytes",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"organisms"
] |
2018
|
Distinct monocyte subset phenotypes in patients with different clinical forms of chronic Chagas disease and seronegative dilated cardiomyopathy
|
Calprotectin is a calcium-binding cytoplasmic protein found in neutrophils and increasingly used as a marker of bowel inflammation . Fecal occult blood ( FOB ) is also a dependable indicator of bowel morbidity . The objective of our study was to determine the applicability of these tests as surrogate markers of Schistosoma mansoni intestinal morbidity before and after treatment with praziquantel ( PZQ ) . 216 children ( ages 3–9 years old ) from Buliisa District in Lake Albert , Uganda were examined and treated with PZQ at baseline in October 2012 with 211 of them re-examined 24 days later for S . mansoni and other soil transmitted helminths ( STH ) . POC calprotectin and FOB assays were performed at both time points on a subset of children . Associations between the test results and infection were analysed by logistic regression . Fecal calprotectin concentrations of 150–300 µg/g were associated with S . mansoni egg patent infection both at baseline and follow up ( OR: 12 . 5 P = 0 . 05; OR: 6 . 8 P = 0 . 02 ) . FOB had a very strong association with baseline anemia ( OR: 9 . 2 P = 0 . 03 ) and medium and high egg intensity schistosomiasis at follow up ( OR: 6 . 6 P = 0 . 03; OR: 51 . 3 P = 0 . 003 ) . Both tests were strongly associated with heavy intensity S . mansoni infections . There was a significant decrease in FOB and calprotectin test positivity after PZQ treatment in those children who had egg patent schistosomiasis at baseline . Both FOB and calprotectin rapid assays were found to correlate positively and strongly with egg patent S . mansoni infection with a positive ameloriation response after PZQ treatment indicative of short term reversion of morbidity . Both tests were appropriate for use in the field with excellent operational performance and reliability . Due to its lower-cost which makes its scale-up of use affordable , FOB could be immediately adopted as a monitoring tool for PC campaigns for efficacy evaluation before and after treatment .
With over 207 million people worldwide infected with schistosomiasis , mostly in sub-Saharan Africa , there is still a pressing need to optimize field-appropriate tools for morbidity staging and monitoring of clinical disease [1] . In recent years there has been a scale-up of schistosomiasis control programs worldwide [2] , [3] . However , no single direct morbidity marker has been adopted to monitor the clinical impact of these interventions [4] . Part of the difficulty is due to the variability of schistosomiasis-related clinical manifestations that are often non-specific and species dependent . This holds particularly true for intestinal schistosomiasis , mainly caused by two different species of the genus Schistosoma: S . japonicum and S . mansoni , in which the clinical diagnostic gold-standard for bowel morbidity is a colonoscopy . This has limited applicability in resource limited settings owing to logistic constraints [5] . Hence it is important to find suitable low-cost reliable proxy markers of bowel morbidity that can aid in measuring the clinical impact of control programs . Intestinal schistosomiasis is an inflammatory disease in which the observable pathology derives from the local inflammatory host response to the egg-entrapment inside the intestinal mucosa and submucosa [5] . It can manifest with a wide range of symptoms depending on the extent of the inflammation and friability of the mucosa , ranging from intermittent abdominal pain to overt acute dysentery . By performing colonoscopy studies , colorectal schistosomiasis has been consistently associated with findings of intestinal polyps and pseudopolyps that can present as blood-per-rectum in affected individuals [6] . Across community-based surveys worldwide , detection of intestinal blood loss in schistosomiasis-endemic areas has revealed a positive association between egg-patent infection and presence of blood in stools [7]–[9] . However , the majority of these studies used guaiac-based fecal occult blood ( FOB ) tests that are known to be less sensitive than the FOB immunochemical assays now widely used for colorectal cancer detection [10] , [11] . FOB has also been used for identifying blood loss in parasitic enteric infections with hookworm , Trichuris trichiura or Entamoeba histolytica [12]–[14] . For intestinal schistosomiasis a longitudinal FOB study in Ugandan children treated with praziquantel ( PZQ ) showed a strong correlation with infection before and one year after intervention , highlighting its potential use as a morbidity marker across field surveys [15] , [16] . Shorter term disease dynamics need investigation , in particular during the standard cure rate monitoring period at 24 days post PZQ treatment [17] . Calprotectin is a neutrophil cytoplasmic calcium-binding protein that is also found in monocytes and early stage macrophages . Its degranulation inside the intestinal lumen occurs as a response to local inflammation [18] . Detection of calprotectin in stool is currently used across gastroenterology practices to aid diagnostically in distinguishing between inflammatory bowel disease and other non-inflammatory ailments , thereby decreasing the number of unnecessary endoscopies performed . It is also used as a validated marker for disease activity and response to treatment [19]–[21] . Its use in enteric infections is gaining recognition , particularly as a correlative marker for clinical severity in infectious diarrhea from both viral and bacterial etiologies [22] . Based on its potential as a surrogate inflammatory maker , we had previously hypothesized that fecal calprotectin could be elevated in schistosomiasis-associated intestinal pathology . However upon field-testing of an enzyme-linked immunosorbent assay , we found no correlation with S . mansoni infection [16] . The study was limited by the complexity and low sensitivity of the ELISA protocol and high minimum calprotectin value detectable in stool . However the low S . mansoni prevalence in the area may have also affected the association with the study outcome [16] . With the recent commercial availability of a new POC calprotectin assay based upon immunochromatographic dipsticks , this new test format has greater sensitisation and detection range of calprotectin found in feces . The primary objective of the present study was an investigation of point-of-care ( POC ) chromatographic FOB detection and fecal calprotectin with S . mansoni infection status as a proxy for intestinal morbidity . Secondary objectives were to ascertain their association with less specific schistosomiasis downstream manifestations such as anemia and to investigate the short-term changes 24 days after PZQ treatment .
Ethical approval was obtained from both the Liverpool School of Tropical Medicine and the Uganda National Council of Science at the Ministry of Health . For all participants information sheets describing the study were delivered before recruitment . Informed written consent was obtained from the children's parents/guardians and verbal assent from the subjects when possible . The study was conducted as an exit investigation from a large longitudinal cohort study based in two villages on the Lake Albert shoreline in Buliisa district , Uganda ( Figure 1 ) during the month of October 2012 with a follow up survey 24 days later in November 2012 . Children had previously been recruited for a longitudinal study targeting young children with schistosomiasis that started in 2009 ( Schistosomiasis in Infants and Mothers ( SIMI ) cohort ) [23] . A total of 216 children were initially surveyed with 211 of them providing stool and urine at follow-up . Only those children with two stool samples at both time points were included in the analysis . Children provided two stool samples on two consecutive days and one urine sample . Duplicate stool smears were prepared from each sample applying the thick Kato-Katz technique for the detection of S . mansoni and other soil-transmitted helminths . [24] . Means of eggs per gram of stool ( epg ) were subsequently calculated from the egg count slide results . Three infection categories according to WHO standards were considered for intestinal schistosomiasis [25]: ( 1–99 epg ) or ‘Light’ , ( 100–399 ) epg or ‘Medium’ and ( >400 epg ) or ‘Heavy’ . A cathodic circulating antigen ( CCA ) assay was also employed to detect the presence of S . mansoni antigen in urine using a commercially available immuno-chromatographic dipstick ( Rapid Medical Diagnostics , Pretoria , South Africa ) [26] . All children after providing stool , were treated with PZQ at standard dosing of 40 mg/kg regardless of parasitological results . All children provided a small finger-prick sample ( 0 . 5 ml ) of blood for estimation of hemoglobin levels by Hemocue ( Ängelholm , Sweden ) . Anemia was then defined according to age-dependent cut-off values [27] . An antigen detection rapid diagnostic test for Pf/non-Pf malaria was also carried out ( Standard Diagnostics , Kyonggi-do , Korea ) . Children found to be positive were treated according to standard national treatment guidelines . A simple chromatographic test was employed for FOB detection ( Mission Test® , Acon Laboratories , San Diego , CA ) , following the manufacturer's instructions . Briefly a small amount of feces was homogenised in a liquid buffer after collection . Two drops of stool suspension were applied to a test cassette and results were visually read after five minutes and categorized as negative ( − ) , trace ( +/− ) and positive ( + ) . A quantitative point-of–care chromatographic immunoassay was used for the detection of fecal calprotectin ( Quantum Blue® , Alpha laboratories , Hampshire , UK ) , according to the manufacturer's instructions . Two drops of homogenised stool were applied inside a plastic cassette that was then inserted into a portable electronic reader that quantified the intensity of the control and test reactions to express a numerical value . The cassette reader was calibrated for each batch of tests . Cut-off values were based on recent clinical consensus [28] . Values of calprotectin >50 µg/g were considered positive [19] . Calprotectin values were further categorised based on findings from clinical gastroenterology studies [20] , [28] . Calprotectin values >160 µg/g are considered within the range of severe intestinal inflammation as seen in inflammatory bowel disease [28] . Guided by this findings , we then created sub-categories to explore the relationship of different inflammatory cut-offs with schistosomiasis; ‘Low’ ( 51–149 µg/g ) , ‘Medium’ ( 150–299 µg/g ) and ‘High’ ( >300 µg/g ) , for subsequent statistical analysis . All data was entered into electronic format using EpiData® ( The EpiData Association Odense , Denmark ) and later analysed using the R statistical package version 2 . 14 . 1 ( The R Foundation for Statistical Computing , Vienna , Austria ) and Microsoft Excel ( Redmond , WA , US ) . Arithmetic and geometric mean of Williams ( GMW ) were calculated for S . mansoni infection intensity . 95% confidence intervals ( CI ) were calculated for percentages by applying the exact method . Primary endpoints of interest were FOB and calprotectin positivity and their association with egg-patent and CCA positive S . mansoni infection status . After initial exploratory bivariate analysis , multivariable logistic regression modelling was carried out using the binomial variables constructed for FOB and calprotectin as the outcome variables and adjusting for possible confounding variables such as malaria , anemia , sex , age and village location . A separate analysis was also performed for different calprotectin intensity values . Odds ratios ( OR ) and 95% CI and P-values were calculated as measures of association between each variable and the outcomes . P-values<0 . 05 were considered statistically significant .
Children were stratified by S . mansoni infection status at baseline ( as assessed by Kato-Katz or CCA test ) and the percentage of children positive for FOB and calprotectin in each group was determined both at baseline and 24 days post PZQ treatment ( Table 2 ) . The results showed that in the group which was egg patent at baseline , there was a significant decrease in numbers of individuals who were FOB or calprotectin positive 24 days after treatment . This was particularly evident for children with medium intensity infections ( for calprotectin P = 0 . 035 ) and heavy intensity infections ( for FOB P = 0 . 007 , and calprotectin P = <0 . 0001 ) . All children with positive FOB were found to be S . mansoni egg positive at baseline and therefore S . mansoni was not included in the baseline analysis . After adjusting for potential confounders ( age , sex , anemia , malaria ) at follow up there was a very strong association between FOB and S . mansoni medium and heavy infection ( OR: 6 . 6 P = 0 . 033 , OR: 51 . 3 P = 0 . 003 ) . Children with anemia were very likely to have a positive FOB test result , but only at baseline ( OR: 9 . 2 , P = 0 . 036 ) and girls were more likely to be FOB positive at follow up ( OR: 3 . 7 , P = 0 . 057 ) . Residents from Walukuba village were more likely to be FOB positive at baseline ( OR: 7 . 5 , P = 0 . 02 ) , however the effect was significantly reversed at follow up ( OR: 0 . 16 , P = 0 . 03 ) . Four different models were constructed after bivariate exploration of the possible associations between fecal calprotectin and S . mansoni infection . When calprotectin positivity ( >50 µg/g ) was considered as a binomial outcome , an association was found at baseline between CCA test positivity , younger age ( 3–5 years old ) and residency in Bugoigo village . These associations , except for the village effect , were not seen 24 days after PZQ treatment . When stratified by calprotectin intensity , ‘Medium’ intensity ( 150–299 µg/g ) strongly correlated with egg-patent S . mansoni infection at baseline ( OR: 12 . 5 , P = 0 . 05 ) and follow up ( OR: 6 . 8 , P = 0 . 02 ) , as well as with FOB positivity at baseline ( OR: 7 . 5 , P = 0 . 05 ) . Before PZQ treatment , younger children ( 3–5 years old ) had significantly increased levels of ‘Light’ intensity ( 51–149 µg/g ) fecal calprotectin ( OR: 2 . 4 , P = 0 . 03 ) as well as increased ‘Medium’ intensity calprotectin at follow up ( OR: 4 . 1 , P = 0 . 03 ) . FOB at follow up was associated with ‘Light’ intensity calprotectin in unadjusted analysis ( Crude OR: 3 . 89 , P = 0 . 05 ) although this was not seen when controlled for potential confounders . Across all intensities , residents of Bugoigo village were more likely to have positive fecal calprotectin .
In the context of performance of PZQ PC in targeted areas for control of morbidity associated with schistosomiasis , there is a growing need for standardized , affordable and non-invasive point-of-care tests as surrogate markers of disease [4] . Our results highlight the feasibility and reliability of both FOB and calprotectin assays as appropriate field-applicable correlates for intestinal schistosomiasis in young children and could be used to monitor morbidity in pre- and post-treatment settings . A positive test may also encourage people to partake in treatment campaigns as there is a generic need for better health advocacy . Egg-patent schistosomiasis is highly prevalent in our study area despite repeated PZQ rounds in the past five years [17] . Indeed transmission here is known to be intense and the regional hot spot of disease for over 30 years [17] . We found that infection significantly correlates with markers of inflammation ( calprotectin ) and mucosal bleeding ( FOB ) and that this had a direct relationship with intensity of infection . Hence we can infer that intestinal schistosomiasis is also highly prevalent in the area and its public health implications remain to be fully appreciated . This is a rather worrisome result for an area with yearly anti-parasitic treatment campaigns and may be attributed at least in part , to the non-inclusion in the programs of pre-school age children that are known to harbour egg-patent infections [29] , [30] . This leaves a large proportion of the infected population untreated despite recent WHO recommendations to include the under-fives in schistosomiasis PC programs [29] , [23] , [31] . Different tools for the detection of intestinal morbidity have been tested in different community surveys with mixed results [4] . In Uganda , a longitudinal study measuring the abdominal circumference ratio ( ACR ) in children with distended abdomens showed the widespread occurrence of a diseased state although with a weak association with morbidity-related egg-patent schistosomiasis [32] . Abdominal ultrasounds are widely used for hepatic and urinary schistosomiasis in areas where S . mansoni and S . haematobium are endemic , but they have little value for intestinal disease detection [33] . Even in a highly endemic S . mansoni area , hepatic and intestinal schistosomiasis may or may not co-exist and this emphasizes the importance of differentiating both clinical entities . As a limitation to our study we could not compare the point-of-care tests that we used with colorectal endoscopy , the gold-standard diagnostic test for intestinal morbidity , however calprotectin has been widely previously validated in other studies as an acceptable inflammatory marker against colonoscopy [19] , [21] , [28] . Our proxy for morbidity was egg-patent S . mansoni infection . Infection is different from clinical disease and particularly in schistosomiasis egg-shedding may decrease with time and treatment , but organ damage from chronic inflammatory response to tissue trapped eggs is likely to persist . The association between FOB and Schistosoma infection has been clearly demonstrated before and it can therefore be considered a good surrogate marker for intestinal pathology [7]–[9] , [15] . The advantages of the FOB immunochemical test are their relatively low cost ( $0 . 6/test ) , easy to use with very little technical expertise required . All tests were reliable with no tests failing making the invalid test rate negligble . The associations between Walukuba village and FOB , as well as between Bugoigo village and calprotection , are intriguing . The precise disease manifestation associated with a parasitic infection depends on a complex interplay between the host and parasite . In recent work , we did not find significant genetic differentiation between S . mansoni parasite populations from the three villages included in the present study ( Betson et al . , under review ) . However , there are tribal differences in the human populations between the three villages with certain tribes more or less represented in each village , for example most people in Walukuba belong to the Alur tribe , whereas in Bugoigo and Piida , the Banyoro and Bugungu tribes are also well-represented . These differences in local ethnography may least in part explain the associations between specific morbidity markers and particular villages . The intertwined relationship between anemia and schistosomiasis is found across species and even in individuals with light parasitic loads [34] , [35] . What remains to be elucidated , however , is the attributable fraction of intestinal losses versus other pathogenic mechanisms such as anemia of inflammation [36] . The burden of anemia in endemic communities leads to poor school and physical performance , such manifestations that will spiral down to an overall decreased quality of life and adult attainment [37] , [38] . We found an interesting positive association between FOB and anemia at baseline , even after controlling for malaria co-infection . This suggests significant amounts of blood loss per rectum which would be consistent with intestinal schistosomiasis with high parasitic loads as reported in S . japonicum endemic settings [7] . The potential confounding effects of other parasitic infections known to cause intestinal bleeding ( i . e . hookworms ) was not of concern in our study site since none were detected . However hookworm should be considered in other endemic areas , e . g . in the Lake Victoria setting [16] . Due to mucosal inflammation and possibly bleeding polyps this presentation would require active clinical intervention [6] . At follow up , we documented a rapid response to PZQ with decreased overall FOB positivity in children with S . mansoni at baseline and no association between FOB and anemia . This is highly suggestive of effective mucosal healing . Fecal calprotectin proved useful as an inflammatory marker in correlation with S . mansoni infection . It is important to note that the calprotectin range most significantly associated with any intensity S . mansoni infection ( 150–299 µg/g ) , is within the inflammatory bowel disease detection range and would warrant further endoscopic work up in industrialized countries [19] , [28] . This is a significant finding since not all intestinal schistosomiasis is thought to be inflammatory in nature and therefore calprotectin detection could aid in identifying more active cases when used as a disease severity marker [5] , [39] . The significant decline in calprotectin levels after PZQ treatment in children with egg patent S . mansoni at baseline , also suggests a positive response to anti-parasitic treatment , strengthening the relationship of the association calprotectin-schistosomiasis . Young children ( <4 years of age ) are known to have higher normal range values of fecal calprotectin . Regardless of this , the association found in our study between young age and high calprotectin levels is above the range of physiological parameters for healthy African children and warrants further investigation [40] . This will help define the inflammation attributable to schistosomiasis in infected children . For now , calprotectin is an expensive POC test ( $10/test ) and its widespread implementation in PC efforts could be restrained as a consequence unless outside international donor support was able to subsidise and reduce substantially the cost per test . Nevertheless , it provides very valuable clinical information and it should be considered useful to monitor intestinal disease , especially within a detailed future research study . For the purpose of this study , we were able to ascertain a strong relationship between FOB , calprotectin and moderate and heavy egg intensities of S . mansoni infection . Even though the number of positive individuals decreased after PZQ treatment , there was no overall statistical significance after 24 days . However , when only children that were S . mansoni egg positive at baseline were analysed , both FOB and calprotectin did have a significant decrease and therefore detect an ameloriation response to PZQ treatment . This indeed suggests a short term mucosal healing response . A longer follow up would be needed to assess longer term responses . e . g . 2–3 month period . What remains to be explored in a larger study , is the correlation between the different stages of mucosal involvement comparing POC assays with endoscopy before and after PZQ administration . This would allow for a clinical staging algorithm for intestinal schistosomiasis that could guide future community based treatment strategies . Due to its low-cost and ease of use , we advocate for the immediate integration of FOB as a morbidity monitoring tool in PC campaigns .
|
The severity of intestinal schistosomiasis , a disease caused by Schistosoma mansoni infection , is likely under-reported in part due to the scarcity of field-appropriate morbidity markers . Downstream potential complications of this disease include anemia , failure to thrive , and chronic multi-organ damage . Point-of-care ( POC ) tools to monitor intestinal schistosomiasis in low resource settings are urgently needed to better quantify the burden of disease in endemic countries and to gauge the clinical impact of scale-up of preventive PC . For the present study in rural Uganda , fecal occult blood and fecal calprotectin were identified as potential surrogate markers of intestinal morbidity . We tested both POC tests and found that they were both associated with active schistosomiasis as detected by eggs in stool with significant decrease in test positivity after PZQ treatment demonstrating short term morbidity reversion . Calprotectin was a strong indicator of intestinal inflammation , however , owing to its high per-test price makes it difficult to scale-up accordingly . Conversely , fecal occult blood was technically feasible , low-cost and had optimal performance as a morbidity marker , hence we strongly advocate for its immediate inclusion as a monitoring tool for PC programs .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2013
|
Fecal Occult Blood and Fecal Calprotectin as Point-of-Care Markers of Intestinal Morbidity in Ugandan Children with Schistosoma mansoni Infection
|
Schistosomiasis and soil-transmitted helminthiasis ( STH ) are widely distributed in Cameroon . Although mass drug administration ( MDA ) of mebendazole is implemented nationwide , treatment with praziquantel was so far limited to the three northern regions and few health districts in the southern part of Cameroon , based on previous mapping conducted 25 years ago . To update the disease distribution map and determine where treatment with praziquantel should be extended , mapping surveys were conducted in three of the seven southern regions of Cameroon , i . e . Centre , East and West . Parasitological surveys were conducted in April–May 2010 in selected schools in all 63 health districts of the three targeted regions , using appropriate research methodologies , i . e . Kato-Katz and urine filtration . The results showed significant variation of schistosomiasis and STH prevalence between schools , villages , districts and regions . Schistosoma mansoni was the most prevalent schistosome species , with an overall prevalence of 5 . 53% , followed by S . haematobium ( 1 . 72% ) and S . guineensis ( 0 . 14% ) . The overall prevalence of schistosomiasis across the three regions was 7 . 31% ( 95% CI: 6 . 86–7 . 77% ) . The prevalence for Ascaris lumbricoides was 11 . 48 ( 95% CI: 10 . 93–12 . 04% ) , Trichuris trichiura 18 . 22% ( 95% CI: 17 . 56–18 . 90% ) and hookworms 1 . 55% ( 95% CI: 1 . 35–1 . 78% ) , with an overall STH prevalence of 24 . 10% ( 95% CI: 23 . 36–24 . 85% ) across the three regions . STH was more prevalent in the East region ( 46 . 57%; 95% CI: 44 . 41–48 . 75% ) in comparison to the Centre ( 25 . 12; 95% CI: 24 . 10–26 . 17% ) and West ( 10 . 49%; 95% CI: 9 . 57–11 . 51% ) regions . In comparison to previous data , the results showed an increase of schistosomiasis transmission in several health districts , whereas there was a significant decline of STH infections . Based on the prevalence data , the continuation of annual or bi-annual MDA for STH is recommended , as well as an extension of praziquantel in identified moderate and high risk communities for schistosomiasis .
Recent years have witnessed an increased interest in the control of neglected tropical diseases ( NTDs ) , and today there exists a global momentum for the control of these diseases , as well as an unprecedented opportunity for cost-effective action , through an integrated control [1]–[5] . Interest in the integrated control of NTDs is currently at an all-time high , due in part to new funding committed by a number of governmental and non-governmental donors , high-level political commitment in the endemic countries , and the existence of donated anthelminthic drugs which can be safely co-administrated and used in a coordinated way to address these scourges [6]–[8] . Four of these diseases are mainly controlled through the ‘preventive chemotherapy’ intervention , i . e . schistosomiasis , soil-transmitted helminthiasis ( STH ) , onchocerciasis and lymphatic filariasis , according to the World Health Organization ( WHO ) recommendations [4] . Schistosomiasis and STH occur throughout the developing world and remain a major public health problem in the poorest communities with enormous consequences for development . Praziquantel is the sole drug for treatment and morbidity control of schistosomiasis in sub-Saharan Africa . Control of STH uses two main drugs , i . e . albendazole or mebendazole . Based on infection prevalence , communities can be classified into low-risk ( <10% for schistosomiasis and <20% for STH ) , moderate-risk ( ≥10% but <50% for schistosomiasis and ≥20% but <50% for STH ) and high-risk ( ≥50% for both ) categories according to the WHO disease specific thresholds , and this classification is used to determine the appropriate treatment regimen as specified in the WHO guidelines [4] . In Cameroon , it is estimated that more than 5 million people are at risk of infection with schistosomiasis , and 2 million persons are currently infected [9] . STHs are widely distributed all over the country , and it is estimated that more than 10 million people are infected with intestinal worms [9] . The national epidemiological survey conducted in 1985–1987 showed the occurrence of three species of schistosomes: Schistosoma haematobium , S . mansoni and S . guineensis ( formerly S . intercalatum Lower Guinea strain [10] , [11] ) ; and three major species of STH: Ascaris lumbricoides , Trichuris trichiura and Necator americanus . The highest transmission levels of schistosomiasis occurred in the savannah areas of the northern Cameroon , whereas STHs were more prevalent in the southern forest part of the country [12]–[14] . School-aged children are the most infected , and polyparasitism is very frequent; with a high proportion of children carrying at least 2 species of parasites [15] . Cameroon adopted a strategic plan for the control of schistosomiasis and STH in 2004 . Starting with very limited budget , the control programme gradually mobilized national and international partners to enable a rapid scaling-up of activities to encompass all ten regions in 2007 . Since then , national deworming campaigns were implemented annually . School-aged children were treated with mebendazole nationwide , whereas praziquantel was distributed only in high endemic areas for schistosomiasis [16] . Interestingly , the Government of Cameroon recently moved into an integrated approach for the control of NTDs , including co-implementation of different control interventions and co-administration of several drugs , i . e . praziquantel , ivermectin , mebendazole and albendazole . This integrated approach is the basis for cost-effectiveness and streamlined efficiency . Since 2009 , Cameroon receives assistance from the United States Agency for International Development ( USAID ) through its NTD Control Program to facilitate integration of national programs and support mass drug administration ( MDA ) [17] . Because knowing the distribution of the targeted NTDs is essential for developing an adequate implementation strategy and types of drug co-administrations , one of the efforts of the USAID's NTD control program in Cameroon was focused on updating the disease-distribution information . Hence , efforts were made to support on-the-ground activities to map the disease distribution where sufficient information was not available . Indeed , the baseline data for schistosomiasis and STH in Cameroon were collected 25 years ago [12] , [13] . It is well known that the transmission of these diseases is dynamic over time , particularly after years of treatment and other health interventions [18] . Therefore , epidemiological surveys were scheduled in the different regions of Cameroon in order to update the distribution and the level of endemicity of schistosomiasis and STH to facilitate the planning of implementation strategies in these regions . The first study phase targeted three of the ten regions of Cameroon , i . e . Centre , East and West . The present paper reports the outcome of the mapping exercises , compares the current situation with the baseline data from 1980s , and provides recommendations for the control of schistosomiasis and STH in these regions .
The study was approved by the National Ethics Committee of Cameroon ( Nr 082/CNE/DNM/09 ) , and was a public health exercise through the Ministry of Public Health and the Ministry of Basic Education . Parasitological surveys were conducted in schools with the approval of the administrative authorities , school inspectors , directors and teachers . Information about the national programme for the control of schistosomiasis and STH , and the objectives of the study were explained to the schoolchildren and to their parents or guardians from whom written informed consent was obtained . Children willing to participate were registered . Each child was assigned an identification number and data collected were entered in a database . No identification of any children can be revealed upon publication . Children were treated during the MDA campaign implemented by the national control programme . Cameroon is divided up into a three-tiered system including 10 regions at the first level , 58 divisions ( departments ) at the second level , and 360 sub-districts ( arrondissements ) at the third level . The population of Cameroon is estimated to be 19 , 406 , 100 inhabitants in 2010 . Population density shows marked variation across the country , ranging from a mean of 7 . 4 inhabitants/km2 in the East region to 141 . 5 inhabitants/km2 in the Littoral region . School-aged children account for 28% of the country population and are estimated at 5 , 433 , 708 [19] . The health system in Cameroon is decentralized and organized into central , regional and district levels . There are 179 health districts . The three regions targeted for mapping , i . e . Centre , West and East are located in the southern forest area of the country . These regions are subdivided in 29 , 14 and 20 health districts , respectively . A stratified random-cluster sampling procedure , with the 5th grade as the basic sampling unit , was used in the previous mapping of schistosomiasis and STH in Cameroon , conducted in 1985–1987 [12] , [20] . In order to assess the current levels of infections and to compare the data with previous ones , the schools were selected using the list of villages and schools previously investigated , the ecological zones and the risk factors for schistosomiasis transmission [21] , [22] . Selection was made so that all health districts in the three targeted regions of Cameroon were covered spatially . Due to financial limitations , an average of four primary schools ( proportional to the district's size and population density ) was selected per health district . The geographical co-ordinates of each of the sampled schools were recorded with global positioning system ( GPS ) devices . The study was conducted in April–May 2010 . In the 1985–1987 study , a 10 ml urine sample and a single Kato-Katz slide were examined for schistosome and STH infections [12] , [20] . In the current study , in each school , urine and stool samples were collected from 50 children selected randomly in the upper classes , approximately half boys and half girls . Children were preferentially selected from the 5th grade , and then in other grades where the number of children in the 5th grade was fewer than 50 . The samples were collected in 60 mL plastic screw-cap vials , between 10 . 00 and 14 . 00 hours . The samples were preserved with sodium azide [12] , [20] and transported to the Centre for Schistosomiasis & Parasitology in Yaoundé for examination . In the laboratory , each urine sample was agitated to ensure adequate dispersal of eggs , 10 mL of urine were filtered through a Nucleopore® filter , and the filters were examined by microscopy for the presence of schistosome eggs . Stool samples were examined by a single thick smear technique using a 41 . 7 mg Kato-Katz template . Each Kato slide was read twice; immediately after slide preparation for hookworm eggs , and the following day for schistosome and other STH eggs . Parasitic infections were recorded; number of eggs for each parasite was counted; and intensity of infection was calculated and expressed as eggs per gram of feces ( epg ) or eggs per 10 ml of urine ( egg/10 ml ) . The different parasitological data were analyzed by the epidemiological unit of the Centre for Schistosomiasis & Parasitology using appropriate statistical tests and methods . The data were subsequently exported into SPSS ( IBM , Version 19 ) for statistical analysis . The Complex Samples Crosstabs procedure was used for calculating the prevalence and the Descriptives procedure was used for calculating the intensity of infections , taking into account the cluster nature of schools with districts as strata and schools as clusters and including the finite population correction assuming equal probability sampling without replacement . Sample weighting was applied for each district according to the ratio of the proportionally expected number of schools to be surveyed and the number of actually surveyed schools in each district assuming similar number of children in each school [23] . The 95% confidence intervals ( CIs ) for prevalence were calculated using the Wilson score method without the continuity correction after adjusting for sample weighting [24] . Arithmetic mean intensities of infection with 95% CIs for different parasite species were calculated including all children examined [25]–[27] . The Chi-square test using the Complex Samples Crosstabs procedure was used to investigate the relationship between prevalence of infections and sex , age groups , districts and regions , and the Complex Samples Logistic regression procedure was used to compare the differences in prevalence between 1985–1987 and 2010 . The Kruskal-Wallis test was used to compare the differences in intensities of infections . The levels of endemicity of schistosomiasis and STH and the degrees of intensity of individual infections were categorized according to the WHO recommendations [4] , [28] . A geographical information system ( GIS ) software ArcGIS ( ESRI Inc . , Version 9 . 2 ) was used to plot the point prevalence of the infections for each surveyed school on a map .
The arithmetic mean intensity of infection in the three regions for each species of schistosomiasis is shown in Table 1 . The egg counts for intestinal schistosomiasis ranged from 0 to 13 , 818 epg , and from 0 to 2 , 600 eggs/10 ml for urinary schistosomiasis . The overall arithmetic mean infection intensity was 33 . 24 epg for S . mansoni , 2 . 46 eggs/10 ml for S . haematobium , and 0 . 23 epg for S . guineensis . The Centre region was most heavily infected with S . mansoni ( 61 . 04 epg ) and the West region with S . haematobium ( 6 . 86 eggs/10 ml ) . It appears that infections were light ( <100 epg ) in the majority of schools , with only 2 . 5% moderate or heavy S . mansoni infections and 0 . 72% heavy S . haematobium infections across the three regions ( Table 1 ) . Boys were more heavily infected with S . mansoni or S . haematobium than girls ( p<0 . 01 ) . The age distribution of intensity of infection for individual schistosome species is shown in Figure 3 . Intensity of infection increased with age for S . haematobium in children examined while children of 9–14 years old were more heavily infected with S . mansoni ( p<0 . 001 ) . The current distribution of schistosomiasis and STH in 2010 was compared with the distribution in 1985–1987 [12]–[14] , using the overall schistosomiasis and STH prevalence . The prevalence distribution of schistosomiasis in 1985–1987 and in 2010 is shown in Figures 5A and 5B , respectively , with the prevalence categorized according to the WHO prevalence thresholds [4] . It shows that the overall endemic areas of schistosomiasis did not change significantly . However , there was an increase in the number of high transmission foci of schistosomiasis in several health districts; e . g . health district of Malantouen in the West region where prevalence was up to 95 . 92% in the village of Matta , and health districts of Mbalmayo and Bafia in the Centre region with prevalence up to 71 . 43% and 52 . 78% in the villages of Dzeng and Yorro , respectively . Statistical comparison was carried out taking into account the geographical location of districts , age and sex . The results are shown in Table 2 . Compared with the 1985–1987 data , the overall schistosomiasis prevalence in 2010 across the three regions and that in the Centre region did not change significantly ( p>0 . 05 ) , while prevalence in the East region decreased and prevalence in the West region increased , both significantly ( p<0 . 01 ) , though the level of infection in these two regions were relatively lower . Among the three schistosome species , the overall S . haematobium prevalence remained unchanged ( p>0 . 05 ) , while the overall S . mansoni prevalence significantly increased from 4 . 3% to 5 . 53% ( p<0 . 05 ) , and that of S . guineensis , though low , decreased significantly ( p<0 . 001 ) . The prevalence distribution of STH in 1985–1987 and in 2010 is shown in Figure 6 . There was a clear and significant decrease of STH prevalence in all three regions . Indeed , statistical comparison showed that the overall STH prevalence declined significantly from 93 . 02% , 92 . 34% and 81 . 14% to 25 . 12% , 46 . 56% and 10 . 51% in the Centre , East and West regions , respectively ( all p<0 . 001 ) ( Table 2 ) . However , the decrease of STH was significantly lower in the East region in comparison to the two other regions . Detailed analysis of individual STH species showed significant reductions of 86 . 99% for hookworms , 82 . 27% for A . lumbricoides and 78% for T . trichiura ( all p<0 . 001 ) . Analysis of polyparasitic infections showed that in 1985–1987 , 61 . 93% of school-aged children examined were infected with more than one and up to five parasite species , but this proportion decreased significantly to 10 . 19% in 2010 ( p<0 . 001 ) ( Table 2 ) .
The present study showed that schistosomiasis was moderately endemic ( prevalence between 10–49% ) in 23 of the 244 schools investigated , and highly endemic ( prevalence ≥50% ) in 4 schools . These moderate and high-risk communities are distributed in 13 of the 63 health districts investigated . The results confirmed the typical focal distribution of schistosomiasis in these regions . When comparing our results with the previous nationwide data collected in 1985–1987 by Ratard et al . [12] , it appears a slight increase of the number of high transmission foci of schistosomiasis and an overall increase of S . mansoni infections – the most prevalent schistosome species in the three regions . This is not surprising given the fact that no MDA with praziquantel had been implemented in these health districts since the last mapping survey , apart from the health district of Ndikinimeki in the Centre region . The national control programme for schistosomiasis and intestinal helminthiasis was officially launched in 2004 in Cameroon [16] . Since 2007 , school-aged children had been dewormed annually with mebendazole nationwide in all 179 health districts , whereas praziquantel were distributed only in schistosomiasis highly endemic health districts , including all 51 health districts of the three northern regions of Cameroon , where the highest transmission level of schistosomiasis were found [12] , [29] , and only in one of the 63 health districts of the three investigated regions , i . e . the district of Ndikinimeki , Centre region . The comparison of 1985 and 2010 data showed a significant decrease of schistosomiasis prevalence within the health district of Ndikinimeki , with a decline from 81 . 60% to 41% in the town of Makenene for example . Changing situation of schistosomiasis varied in the three regions and among the three different species , and this may reflect the differences in transmission dynamics in these different regions . The main factors influencing schistosomiasis transmission may include the changing demographic situation , socioeconomic development , water and sanitation , snail population dynamics etc . However , such information was not collected in the current mapping survey , which may be a topic for future studies . One of the key outcomes and recommendations from this study is that in future deworming campaigns , the distribution of praziquantel should be undertaken in all 13 health districts in these three regions where schistosomiasis prevalence were ≥10% , according to WHO preventive chemotherapy guidelines [4] . Considering the overall low endemicity of schistosomiasis in the majority of these health districts , treatment will be conducted at district level in rural zones , whereas in urban settings treatment will be focused in those sub-districts with high prevalence spots of schistosomiasis . It should be noted that in both 1985 and 2010 surveys , single Kato-Katz slides were conducted as commonly used for mapping studies . Therefore , the prevalence and abundance of S . mansoni and STH may have been underestimated , due to the low sensitivity of Kato-Katz technique and day-to-day variation in egg excretions , particularly in light infections . For STH , our study showed an overall significant decrease of infection prevalence in all three regions investigated , in comparison to previous mapping data collected in 1985–1987 [13] , [14] , [29] . Indeed , the STH prevalence declined from 93% to 25 . 1% in the Centre region , from 81% to 10 . 5% in the West region , and from 92 . 3% to 46 . 6% in the East region . These results clearly illustrate the positive impact of the school-based deworming campaigns with mebendazole implemented annually by the Ministry of Public Health , through the National Programme for the Control of Schistosomiasis and Intestinal Helminthiasis . The decline was lower in the East region compared to the two other regions . The previous mapping data showed that the three regions surveyed were among the higher STH prevalence areas within the country [13] , [14] . Apart from the ivermectin MDA implemented in onchocerchiasis endemic communities , these regions have not been subjected to albendazole distribution which is used for lymphatic filariasis control . It is therefore interesting to see that the overall STH prevalence has been reduced so much by mainly mebendazole distribution . Though it has been shown that mebendazole is not as efficient as albendazole in deworming , particularly for hookworms [30]–[32] , the present data show that mebendazole still has a significant role to play in the current effort to control NTDs . Several other factors , such as socio-economic development , improved sanitation and hygiene , environmental changes and collateral effect of other drugs , may have also contributed to the reduction of STH transmission . However , as discussed above this may be a topic for future studies . Despite the observed significant reduction of STH infections , the prevalence and intensities of A . lumbricoides and T . trichiura infections were still relatively high , particularly in the East region . Several factors may explain the lower reduction of STH infections in this region , including the low socio-economic status and poor sanitation in most of the rural settings , which favor high parasite transmission and frequent human re-infections . The East region is the largest and the most sparsely populated region in Cameroon . The vast majority of its inhabitants being subsistence farmers , the low level of development in the region , and its thick forests and equatorial climate are favorable factors for STH and other NTDs . Also , the lower school attendance rates in villages , in comparison to towns , may have affected the treatment coverage of all school-aged children through a single school-based deworming campaign approach . It is well known that the epidemiology of STH infections is influenced by several determinants , including environment , population heterogeneity , age , household clustering , genetics and polyparasitism [33] . STHs affect the poor and infections are particularly abundant among people living in rural or deprived urban settings with low socio-economic status and poor sanitation [34] . Further investigations should be conducted to identify the major factors affecting the deworming effect in order to improve the impact of the current integrated NTD control programme . The mapping results showed that the majority of health districts ( 34 over the total of 63 , i . e . 53 . 97% ) were still within the STH infection categories requiring large-scale preventive chemotherapy interventions , i . e . infection prevalence ≥20% . In communities with prevalence ≥50% , WHO recommends treatment of all school-aged children – enrolled and not enrolled – twice per year , and even three times if resources are available; whereas in communities where prevalence is ≥20% but <50% , school-aged children should be treated once a year . Therefore , the government of Cameroon should continue implementing annual deworming of school-aged children in all districts of the Centre , East and West regions . In addition , preschool children , women of childbearing age and adults at high-risk in certain occupations should also be treated , according to WHO recommendations [4] . In particular , in the East region where STH infection prevalence and intensities remain very high , it should be envisaged to deworm school-aged children at least twice a year . Furthermore , the alternating use of mebendazole and albendazole from one deworming round to another should be envisaged to optimize treatment efficacy against STHs [35] . Finally , the results of the present study highlight the new health districts where the MDA of praziquantel should be implemented for the treatment of schistosomiasis . For future deworming campaigns , all school-aged children should be treated with praziquantel in moderate ( i . e . prevalence ≥10% but <50% ) and high-risk communities ( i . e . prevalence ≥50% ) . Also , praziquantel should be made available in dispensaries and clinics for treatment of suspected cases , in accordance with WHO recommendations [4] . Interestingly , this study provided data for accurate estimation of increased praziquantel needs , and the results will contribute to update global information on the distribution of schistosomiasis and STH , recently developed as an open-access database [36] , [37] .
|
Schistosomiasis and soil-transmitted helminthiasis ( STH ) are a major public health problem in Cameroon . The national control strategy of these diseases was based on historical data collected 25 years ago , which might be outdated in some situations due to several factors including control activities , improved or degraded sanitation and hygiene , socio-economic improvement and disease transmission dynamics . To help planning , improving control strategies and evaluation of control activities , there was a need to update the distribution of schistosomiasis and STH . We conducted parasitological surveys in three regions of Cameroon , i . e . Centre , East and West . Our results showed a significant decrease of STH infection prevalence and intensities in all these three regions , in comparison to previous mapping data , with an overall decline of prevalence from 81 . 1–93% to 10 . 5–46 . 6% . These results show the positive impact of annual deworming campaigns , and illustrate the progressive success of the national programme for the control of schistosomiasis and STH in Cameroon . Furthermore , our results showed an increase of the number of high transmission foci of schistosomiasis , and allowed identifying new health districts requiring mass treatment with praziquantel , and those where deworming should be reinforced .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"neglected",
"tropical",
"diseases",
"parasitic",
"diseases"
] |
2012
|
Mapping of Schistosomiasis and Soil-Transmitted Helminthiasis in the Regions of Centre, East and West Cameroon
|
Virus infection of mammalian cells induces the production of high levels of type I interferons ( IFNα and β ) , cytokines that orchestrate antiviral innate and adaptive immunity . Previous studies have shown that only a fraction of the infected cells produce IFN . However , the mechanisms responsible for this stochastic expression are poorly understood . Here we report an in depth analysis of IFN-expressing and non-expressing mouse cells infected with Sendai virus . Mouse embryonic fibroblasts in which an internal ribosome entry site/yellow fluorescent protein gene was inserted downstream from the endogenous IFNβ gene were used to distinguish between the two cell types , and they were isolated from each other using fluorescence-activated cell sorting methods . Analysis of the separated cells revealed that stochastic IFNβ expression is a consequence of cell-to-cell variability in the levels and/or activities of limiting components at every level of the virus induction process , ranging from viral replication and expression , to the sensing of viral RNA by host factors , to activation of the signaling pathway , to the levels of activated transcription factors . We propose that this highly complex stochastic IFNβ gene expression evolved to optimize both the level and distribution of type I IFNs in response to virus infection .
Eukaryotic cells respond to extracellular signals and environmental stresses by coordinately activating specific sets of genes . Signals from the cell surface or cytoplasm trigger signaling pathways that culminate in the binding of distinct combinations of coordinately activated transcription factors to promoter and enhancer elements that regulate gene expression . A well-characterized example of this is the activation of type I interferon ( IFN ) gene expression in response to virus infection or double-stranded RNA ( dsRNA ) treatment [1] , [2] . After infection , viral RNA is detected in the cytoplasm by one of two RNA helicases , retinoic acid-inducible gene I ( RIG-I ) or melanoma differentiation-associated gene 5 ( MDA5 ) , which respond to different types of viruses [3] . RIG-I recognizes short dsRNA or panhandle RNA bearing a 5′ triphosphate group [3] , and its activity is positively regulated by the ubiquitin E3 ligase tripartite motif 25 ( Trim25 ) [4] . When RIG-I or MDA5 bind to RNA , they form heterodimers , undergo a conformational change , and expose a critical N-terminal caspase-recruiting domain ( CARD ) [5] , [6] . This domain interacts with the CARD domain of the downstream adaptor protein mitochondrial antiviral signaling ( MAVS ) ( also known as IPS-1/Cardif/VISA ) on the mitochondrial membrane [7] . The association of RIG-I with MAVS initiates the recruitment of adaptor proteins and leads to the activation of the transcription factors IFN regulatory factors 3 and 7 ( IRF3 and IRF7 ) and NF-κB by the TANK-binding kinase 1 ( TBK1 ) [8]–[10] and IKKα and IKKβ , respectively [7] , [11] . Activated IRF3/IRF7 and NF-κB translocate into the nucleus and , along with the transcription factors ATF2/cJun , bind the IFN-β gene enhancer and recruit additional transcription components to form an enhanceosome [12] . This complex signaling and promoter recognition mechanism functions to coordinately activate a specific set of transcription factors that recognize the unique enhancer sequence of the IFNβ gene and thus specifically activate IFN gene expression . Early in situ hybridization ( ISH ) studies revealed that induction of IFNβ expression by virus infection or dsRNA treatment in both human and mouse cells is stochastic [13] , [14] . That is , only a fraction of the infected cells express IFNβ . This “noisy” expression is not due to genetic variation within the cell population , as multiple subclones of individual cells display the same low percentage of cells expressing IFNβ [14] . In addition , different mouse and human cell lines display different percentages of expressed cells , and the levels of IFNβ gene expression can be increased in low expressing cell lines by fusing them with high expressing lines , or by treating low expressing lines with IFNβ [13] , [14] . These studies suggest that stochastic expression of the IFNβ gene is a consequence of cell-to-cell differences in limiting cellular components required for IFN induction , and that one or more of the limiting factors are inducible by IFNβ [13] . Stochastic expression has been observed with a number of other cytokine genes , including IL-2 [15] , IL-4 [16] , [17] , IL-10 [18] , IL-5 , and IL-13 [19] . In many of these cases , expression is both stochastic and monoallelic . Recent studies of IFNβ gene expression revealed that stochastic expression in human cells is initially monoallelic , and becomes biallelic later in the induction [20] , [21] . In one study the stochastic expression of the IFNβ gene was proposed to be a consequence of intrinsic noise due to stochastic enhanceosome assembly [21] . Subsequently , an analysis of human HeLa cells identified a specific set of Alu-repetitive DNA sequences bearing NF-κB binding sites that associate with the IFNβ gene through interchromosomal interactions , and in so doing are thought to increase the local concentration of NF-κB . Initially , only one of the two chromosomes associates with the specialized NF-κB binding sequence , resulting in early monoallelic expression . Secretion of IFN leads to an increased expression of limiting factors ( most likely IRF7 , which is inducible by IFN ) , obviating the need for interchromosomal interactions , and leading to the activation of the second IFNβ allele [20] . More recently , heterogeneity in the infecting viruses , rather than cell cycle differences , has been proposed to be the primary source of IFN stochastic expression [22] . Many functions have been proposed for biological noise , ranging from cell fate decisions during development to survival in fluctuating environments [23] . In the case of the IFN genes , neither the mechanisms nor functions of biological noise are well understood . Here we report a detailed analysis of stochastic IFNβ gene expression in mouse cells . We make use of an IFN-IRES-YFP reporter mouse [24] to perform a detailed analysis of differences between virus-infected cells that either express or do not express IFNβ . Our results reveal a complex picture of stochastic expression of the IFNβ gene , in which the levels of components required for virtually every step in the virus induction pathway are limiting . This includes components required for viral replication and expression , for sensing the presence of viral RNA by the host , and for the virus induction signaling pathway , and the transcription factors required of IFNβ gene expression . Remarkably , in spite of this complexity the percentage of expressing cells remains constant through recloning and cell division , indicating that the stochasm of clonal cells is genetically programmed .
Sendai virus ( SeV ) infection of either mouse or human cells leads to the expression of IFNβ mRNA in only a fraction of the infected cells ( Figures 1A , 1B , and S1A ) , and the percentage of expressing cells differs between different cell lines . The time course of mouse IFNβ expression determined by ISH ( Figure 1B ) is consistent with that from the quantitative PCR ( qPCR ) analysis ( Figure S1B and S1C ) . Remarkably , the percentage of cells expressing IFN did not exceed 20% , even at the latest time point ( Figure 1B ) . The absence of IFNβ signal in the majority of cells is not an artifact of hybridization , as β-actin mRNA was detected in all cells ( Figure S1D ) . IFNβ mRNA is specifically detected with an antisense IFNβ RNA probe , while no signal is detected with a sense RNA probe ( Figure S1E ) . In addition , similar percentages of IFNβ-expressing cells were detected by immunofluorescent staining using an IFNβ antibody ( Figure S1F ) , strongly supporting the reproducibility and specificity of the IFNβ ISH . As mentioned above , enhanceosome assembly and limiting amounts of NF-κB have been proposed to be the primary limiting steps in stochastic expression of the human IFNβ gene [20] , [21] . To determine whether this stochastic expression is unique to the IFNβ gene because of the complexity of the IFNβ enhanceosome , or is more general , we examined the expression of the IFNα genes , which are coinduced with IFNβ , but have simple enhancer/promoters , and do not require NF-κB [25] , [26] . Using either a mouse IFNα4 or human IFNα8 probe , we found that IFNα genes are also stochastically expressed in both mouse and human cells , respectively ( Figures 1C and S1G ) . Although NF-κB has been shown to be a limiting factor in the activation of the human IFNβ gene [20] , it is not required for IFNβ expression in mouse cells [27] . Thus , in spite of this difference both the mouse and human IFNβ genes are stochastically expressed . We also examined other virus-inducible genes , and found that they too are stochastically expressed ( see below ) . Each of these virus-inducible genes requires different levels and combinations of transcription factors , yet they are all stochastic . In all of these cases ( mouse and human IFNβ and IFNα and the other virus-inducible genes ) , the common requirement is the RIG-I virus-inducible signaling pathway . We therefore carried out experiments to determine whether limiting components in this pathway contribute to the observed stochastic expression . To investigate the mechanism of stochastic IFNβ gene expression , we made use of an IFNβ reporter-knock-in mouse , in which YFP expression allows tracking of IFNβ expression at a single-cell level [24] . Using IFNβ/YFP homozygous mouse embryonic fibroblasts ( MEFs ) and fluorescence-activated cell sorting ( FACS ) , we obtained pure populations of IFNβ-producing and IFNβ-negative cells upon SeV infection . As expected , IFNβ mRNA is high in the YFP-positive cells , and very low in the YFP-negative cells ( Figure S2A ) . As expected , the IFNα2 and IFNα4 genes are also highly expressed in the YFP-positive cells , and not in the YFP-negative cells ( Figure S2A ) . These observations indicate that replication of the infecting virus and/or components in the RIG-I pathway are the limiting steps in the uninduced cells , rather than intrinsic differences in the IFNβ and α promoters . We also detected the relative mRNA abundance of other virus-inducible genes in IFNβ-expressing and non-expressing cells . As shown in Figure 1D , transcription levels of all tested inflammatory cytokine or chemokine genes are much higher in IFNβ-producing cells compared to nonproducers . Considering the fact that IFNβ-producing cells account for only 10% of the total cell population , we conclude that expression of all these virus-inducible genes is also stochastic and that these genes are coordinately activated with the type I IFN genes . Activation of these virus-inducible genes is known to require the RIG-I signaling pathway [28]–[31] . Thus , our results indicate that stochastic gene expression is due primarily to limiting components in the signaling pathway and not to gene-to-gene variation in the mechanism of gene activation . In the case of human cells , stochastic expression of the IFNβ gene is randomly monoallelic early and biallelic late in infection , and the activation of the second IFNβ allele is inducible by IFN [20] , [21] . However , the nature of allelic expression of the IFNβ gene has not been addressed in mouse cells . By using IFNβ/YFP heterozygous MEFs , we showed that early after infection ( <8 h post-infection [h . p . i . ] ) , IFNβ gene expression was primarily monoallelic , while late in infection ( 8–16 h . p . i . ) , the majority of IFNβ-expressing cells were both IFNβ and YFP double-positive cells indicating that , as with human cells , a switch to biallelic expression also occurs in mouse cells ( Figure S2B ) . Previous studies have shown that the levels of IFNβ gene expression can be increased by priming the cells with IFNβ [13] . Using both mouse and human primary fibroblasts , we showed that IFNβ pretreatment also increases the percentages of IFNβ-expressing cells ( Figure S3 ) , indicating that the limiting factor ( s ) contributing to stochastic IFNβ gene expression are , indeed , inducible by IFNβ . One of these IFN-inducible factors is IRF7 ( [20] and see below ) . To examine the role of the infecting virus in stochastic IFNβ gene expression , we infected primary MEFs with SeV followed by immunofluorescent staining using a SeV antibody . As shown in Figure S4A , most , if not all , of the cells are uniformly infected by SeV , far more than could explain the small percentage of cells expressing IFNβ gene . When we used increasing multiplicities of SeV ( as defined by hemagglutination units [HAU] ) to infect primary MEFs , we found that the percentage of IFNβ-producing cells increased as the HAU was increased , reaching a maximum of approximately 18% at the peak ( Figure S4B ) . However , as more virus was added ( >200 HAU ) , the percentage of IFNβ-producing cells decreased . Thus , the viral titer is not a limiting factor in the observed stochastic IFNβ gene expression . Next , we determined viral transcript levels in both IFNβ-producing and nonproducing cells . We found that the nucleoprotein ( NP ) , matrix protein , and L polymerase protein mRNA transcripts were present at significantly higher levels in IFNβ-producing cells compared to the nonproducers ( Figures 2A and S4C ) . In addition , higher levels of SeV NP protein were detected in IFNβ-producing cells ( Figure 2B ) . The RNA helicase RIG-I detects viral genomic RNA and defective interfering ( DI ) genomes [32] , [33] . We therefore examined the levels of viral and DI genomes in both IFNβ-producing and nonproducing cells . As shown in Figure 2C ( upper panel ) , more SeV DI genomes were detected in IFNβ-producing cells compared to IFNβ-nonproducing cells at 8 and 12 h . p . i . Using a primer pair that specifically detects viral genomic RNA , we also detected more viral genomes in IFNβ-producing MEFs 8 and 12 h . p . i . ( Figure 2C , lower panel ) . These results are consistent with the observed viral NP mRNA levels ( Figure S4C ) , and indicate that viral replication is more efficient in the IFN-producing cells . We also investigated the induction activities of total RNA extracted from both IFNβ-producing and nonproducing cells . As shown in Figure S4D , total RNA from IFNβ-producing cells infected for 8 or 12 h induced more IFNβ expression compared to total RNA from IFNβ-nonproducers at the same time points . We conclude that viral mRNA , DI genomes , and viral genomes are present at higher levels in IFNβ-producing cells than in nonproducers . Thus , differences in the efficiency of viral replication/transcription contribute to the stochastic expression of the IFNβ gene . Previous studies led to the conclusion that the stochastic expression of the IFNβ gene is a feature of the infecting virus , and not of the host cell [22] . To address this possibility , we determined the number of cells that have high levels of viral RNA and produce IFNβ at 8 h . p . i . As shown in Figure 2D , after 8 h of virus infection , approximately 38% SeV-high cells ( upper left and upper right ) were detected , and about 9% YFP-positive cells ( upper right and lower right ) . Although a higher percentage of IFNβ-expressing cells was observed within the SeV-high cell population ( 6 . 56% versus 2 . 42% ) , only 17% ( 6 . 56% out of 38% ) of SeV-high cells produce IFN . Thus , although cell-to-cell differences in viral replication contribute to the stochastic expression of IFN , these differences are not sufficient to explain the extent of stochastic IFN gene expression . To further investigate the mechanism of stochastic IFNβ gene expression , we determined the localization of various components of the signaling pathway required for IFN production using nuclear and cytoplasmic fractions separated from both expressing and non-expressing cells . Consistent with the limiting component hypothesis , we detected phosphorylation and translocation of IRF3 in the YFP-positive cells , but not in the YFP-negative cells ( Figure 3A ) . Previous studies have shown that IRF3 , like IRF7 , is phosphorylated by the TBK1 kinase , and translocates from the cytoplasm to the nucleus . As both IRF3 and IRF7 are activated via the RIG-I pathway , our results suggest that one or more components of the RIG-I signaling pathway are limiting in the cells that fail to express IFN . A similar result was obtained with sorted cells at 12 h . p . i . ( Figure 3B ) . In human cells both NF-κB and IRF3/IRF7 are required for virus induction of the IFNβ gene [12] , [34] . The human and mouse IFNβ enhancers differ in only two nucleotides out of 45 bases . However , in mouse cells NF-κB is required only for early antiviral activity , when the level of active IRF3 is low , but is not required for maximum levels of IFNβ expression late in induction [27] , [35] . Consistent with this finding , we show that only a small fraction of the p65 subunit of NF-κB translocates to the nucleus 8 h . p . i . , and little difference is observed in NF-κB localization between the YFP-positive and YFP-negative cells ( Figure 3A ) . The observation that IRF3 activation and translocation occurs in only a fraction of virus-infected cells suggests that upstream components in the RIG-I signaling pathway differ in IFNβ-producing and nonproducing cells . Western blotting results ( Figure 3C ) showed that IFNβ-producing cells have higher levels of both RIG-I and MDA5 than the nonproducing population . Trim25 , an E3 ligase required for RIG-I activation [4] , is also present at a higher level in the IFNβ-producing cells ( Figure 3C ) . The increase in protein levels appears to be a consequence of differential transcription of the tested genes , as mRNA levels of all three genes are higher in IFNβ-producing cells ( Figure 3D ) . We conclude that the IFNβ-producing cells have higher levels of essential RIG-I signaling pathway components than the IFNβ-nonproducing cells . Thus , at least part of the observed stochastic expression is due to limiting RIG-I pathway components in the cells that do not express IFN . By contrast to the RNA detectors , the protein levels for both MAVS and TBK1 , two essential components of the RIG-I signaling pathway [7] , [9] , were lower in the IFNβ-producing cells ( Figure 3C ) . However , this is likely due to the degradation and/or cleavage of the MAVS protein in infected cells [36]–[38] . The data of Figure 3C suggest that TBK1 is also targeted for degradation during virus infection , consistent with the observation that TBK1 is subject to proteasome-dependent degradation [39] . Thus the turnover of both MAVS and TBK1 may be required for the post-induction turn-off of IFNβ gene expression [38] . We have shown that the RIG-I signaling pathway is selectively activated in IFNβ-expressing cells , and this is due only in part to the cell-to-cell differences in virus infection/replication . Our results also suggest that IFNβ-producing cells have a more potent signaling pathway than IFNβ-non-expressing cells . To further explore this possibility , we established a series of L929 stable cell lines that express RIG-I , MDA5 , or Trim25 under the control of a tetracycline-inducible promoter ( Figure S5A ) . As shown in Figure S5B and S5C , high levels of exogenous RIG-I only slightly increased the percentage of IFNβ-producing cells . A larger increase was observed with MDA5 and Trim25 , but the final percentage in both cases was still under 30% . Thus , these upstream components appear to be among several limiting factors in the cell population . Additional components in the RIG-I signaling pathway were tested using the same approach , and high percentages of IFNβ-producing cells were observed ( Figure 4A and 4B ) . While a large difference between tetracycline-negative and tetracycline-positive cells was observed with the TBK1 line , only a small difference was observed between the corresponding MAVS lines . However , a large difference was observed between the non-transformed and transformed MAVS lines , suggesting that a low level of leaky transcription in the MAVS line is sufficient to dramatically increase the number of IFNβ-expressing cells . These data clearly indicate that both MAVS and TBK1 are limiting components in the RIG-I pathway and therefore contribute significantly to stochastic IFNβ expression . We have shown that over-expression of RIG-I or Trim25 alone only slightly increases the percentage of IFNβ-producing cells , but it is possible that both must be expressed to achieve maximum levels of IFNβ production . We therefore transfected RIG-I stable transfectants with a Trim25 expression plasmid , and the other way around . The cells were then induced with tetracycline , infected with SeV , and examined for IFNβ mRNA expression . Control experiments using a GFP reporter indicated that under our experimental conditions approximately 70% of cells can be transfected with the second plasmid ( Figure S5D ) . As shown in Figure 4C and 4D , a dramatic increase was observed only 6 h . p . i . when either the RIG-I or Trim25 lines were transfected with Trim25 or RIG-I , respectively . This observation was confirmed by carrying out intracellular staining and flow cytometry experiments using IFNβ/YFP homozygous MEFs ( Figure S6 ) . We conclude that the combination of RIG-I and Trim25 is limiting in the RIG-I pathway . We note that the increase of IFNβ-expressing cells was not observed in uninfected cells , with the only exception being MAVS . Thus , over-expression of these signaling components did not bypass the requirement for signaling pathway activation . Expression of the IFNβ gene requires an active RIG-I signaling pathway and assembly of the enhanceosome complex on the IFNβ promoter . To investigate whether individual enhanceosome components are limiting factors , we established a series of tetracycline-inducible L929 stable lines that express IRF3 , IRF7 , or p65 genes . Figure 5A and 5B show that , without tetracycline induction , only 10%–15% of the cells produce detectable levels of IFNβ mRNA in response to virus infection . Remarkably , the percentage of IFNβ-producing cells upon SeV infection increased to 85% when IRF7 expression was induced by tetracycline in every cell ( Figure S7A ) . A smaller increase ( 55% ) was observed when IRF3 was over-expressed , whereas increasing the concentration of NF-κB had little effect , consistent with the data in Figure 3A , and previously published studies [27] . Interestingly , IRF7 over-expression also significantly increased the percentage of IFNα-producing cells after virus infection ( Figure S7B and S7C ) . It is known that IRF7 is required for maximum induction of type I IFN genes [25] , and its basal protein level is very low in most cell types except for plasmacytoid dendritic cells [26] , [40] . We conclude that IRF7 is a critical limiting factor that is a major contributor to stochastic expression of mouse IFNα and β genes . This conclusion is also supported by our ISH results from 4E-BP1/4E-BP2 double-knockout MEFs ( Figure 5C and 5D ) . Previous studies have indentified 4E-BPs as negative regulators of type I IFN production via translational repression of IRF7 mRNA [41] . As shown in Figure 5C and 5D , we observed a 4-fold increase of the percentage of IFNβ-expressing cells in 4E-BP1/4E-BP2 double-knockout MEFs compared to wild-type MEFs , consistent with the conclusion that a limiting amount of IRF7 is a major contributor to the stochastic expression of IFNβ . We also found that type I IFN induction was exceptionally high , with much faster kinetics in cells expressing exogenous IRF7 than in control cells ( Figure S7D ) . In the absence of tetracycline induction , low levels of IFNβ , IFNα4 , and IFNα2 mRNA were first detected 6 h , 9 h , and 12 h . p . i . , respectively . When the cells were treated with tetracycline , the kinetics of IFN gene transcription changed significantly . IFNβ , IFNα4 , and IFNα2 transcripts could be detected as early as 4 h after virus infection . Even at 24 h . p . i . , steady and robust transcription of these genes could still be detected . These observations are consistent with a model in which IRF3 is normally activated early for IFN gene induction . Later , higher levels of IRF7 are produced by IFN and are required for both IFNβ and IFNα gene expression , but IRF7 is rapidly turned over , leading to the cessation of both IFNβ and IFNα gene expression [1] , [25] , [26] . By contrast , in the presence of excess IRF7 in the tetracyline-activated cells , both IFNβ and IFNα are activated earlier , and continue to be expressed because of the continuous presence of IRF7 . We have shown that over-expression of IRF7 or both RIG-I and Trim25 almost completely eliminates stochastic IFNβ expression ( Figures 4C , 4D , and 5 ) . To investigate the connection between these observations , we carried out microarray analysis to compare genome-wide expression profiles of L929-IRF7 stable transfectants treated with or without tetracycline . Interestingly , upon IRF7 over-expression , only two up-regulated signaling pathways were identified from the KEGG Pathway Database , and the RIG-I-like receptor signaling pathway is the most up-regulated ( p = 3 . 6E-06 ) ( Figure S8A and S8B ) [42] . We did not identify signaling pathways that were similarly enriched among the down-regulated genes . Using qPCR , we confirmed that the mRNA levels of both RIG-I and Trim25 were higher in IRF7 over-expressing cells ( Figure S8C ) . Considering the low basal expression level of IRF7 , we conclude that a high level of IRF7 protein increases the percentage of IFNβ-expressing cells not only by increasing its own abundance , but also by up-regulating the RIG-I signaling pathway to increase the potency of activation of the IFNβ gene . IFNβ gene expression can also be induced by transfection of the synthetic dsRNA polyriboinosinic polyribocytidylic acid ( poly I∶C ) , and this induction occurs mainly through the MDA5 signaling pathway [43] . Early studies revealed that induction of IFNβ expression by dsRNA treatment is also stochastic [13] , [14] . We therefore asked whether stochastic IFNβ gene expression induced by dsRNA is due to cell-to-cell variation in the levels of MDA5 and IRF7 . Using FACS analysis , we found that poly I∶C–induced IFNβ expression is also stochastic ( Figure 6A ) . When IFNβ/YFP homozygous MEFs were electroporated with Cy5-labeled poly I∶C , only 9% of the cells produced IFNβ as detected by the presence of YFP . However , the electroporation efficiency was over 99% ( Figure 6A , left panel ) . Interestingly , based on the Cy5 intensity , there were two populations of cells , which contained different amounts of poly I∶C . When we gated these two populations out as “poly I∶C–high” and “poly I∶C–low” , we observed that the “poly I∶C–high” population included more cells producing IFNβ ( Figure 6A , right panel ) , indicating that the amount of inducer does affect the extent of stochastic IFNβ expression . However , only a small percentage of “poly I∶C–high” cells expressed the IFNβ gene , clearly indicating that other limiting factor ( s ) dominate the stochastic IFNβ expression induced by poly I∶C transfection . We therefore carried out experiments to identify these limiting components . L929-MDA5 and L929-RIG-I stable transfectants were transfected with poly I∶C followed by ISH to detect IFNβ expression . As shown in Figure 6B and 6C , over-expression of RIG-I only slightly increased the percentage of IFNβ-producing cells . By contrast over-expression of MDA5 , the major cytoplasmic receptor for poly I∶C , led to a substantial increase in the percentage of IFNβ-producing cells ( from 15% to 65% ) . Considering that the transfection efficiency is approximately 75% ( data not shown ) , over-expression of MDA5 basically eliminates stochastic expression of the IFNβ gene in response to poly I∶C transfection . Furthermore , the results of the flow cytometry experiment also supported this conclusion . As shown in Figure 6D , after 8 h of poly I∶C stimulation , we observed approximately 2 . 6% YFP-positive cells . Within this population , about 70% of the YFP-positive cells had higher levels of MDA5 protein ( 1 . 86% out of 2 . 67% ) . We note that the percentage of YFP-positive cells is much lower than that observed with virus infection ( Figures 2D and S6 ) . Over-expression of IRF3 or IRF7 also increased the percentage of IFNβ-producing cells in response to poly I∶C ( Figure S9A and S9B ) . As shown in Figure S8 , over-expression of the IRF7 gene up-regulates MDA5 gene expression . Considering its low basal expression level , IRF7 is also an important limiting factor in stochastic IFNβ expression induced by poly I∶C transfection . Taken together , these data show that poly I∶C–induced stochastic IFNβ expression depends on the abundance of both poly I∶C and signaling pathway protein MDA5 as well as IRF3/IRF7 , which is similar to what was found in the case of virus infection . We also asked whether the concentrations of proteins regulating IFNβ expression are sufficiently different from cell to cell to account for the stochastic IFNβ expression . Using flow cytometry , we measured the distributions of six components in the RIG-I signaling pathway for which specific antibodies are available . As shown in Figure 7A and 7B , all six proteins were log-normally distributed across the population . Quantitative immunofluorescence data for individual components show similar distributions of each factor at the single-cell level ( Figure S10 ) . Combined with our previous data , these observations suggest that naturally occurring differences in the protein levels of signaling pathway components are the primary cause of cell-to-cell variability in IFNβ expression upon virus infection . When IFN is secreted from virus-infected cells in vivo , it binds to type I IFN receptors on surrounding cells and activates a large set of genes encoding antiviral proteins ( interferon-stimulated genes [ISGs] ) via the Jak/STAT signal transduction pathway . We therefore carried out experiments to determine whether the induction of antiviral ISGs is also stochastic . As shown in Figure 7C , ISG15 is expressed in all cells upon treatment with IFNβ . Thus , when IFN is secreted , all of the surrounding cells produce antiviral proteins . This result is also consistent with previous observations showing that the antiviral response induced by IFN is a robust feature common to all cells , and is independent of the stochastic expression of IFN receptor IFNAR [44] .
Regulation of type I IFN production is essential for the innate immune response to viral infections [45] , [46] . However , high levels of IFNβ can be toxic [47] , [48] . Thus , IFNβ production must be tightly regulated . This regulation appears to be both temporal and stochastic . Type I IFN genes are tightly repressed prior to virus infection , activated upon infection , and then rapidly turned off several hours later ( Figure S1B and S1C ) . Previous studies of several cytokine genes suggest that this stochastic gene expression provides an additional mechanism of regulation whereby optimal levels of cytokine production are determined by the frequency of expressing cells rather than by protein levels per cell [18] , [19] , [49] . Thus , it is possible that stochastic expression is a primary mechanism for controlling the optimal level of IFNβ production in vivo . In particular , we have shown that while IFN production is stochastic , the activation of the antiviral gene program by secreted IFN is not . Thus , stochastic expression of IFN would allow the regional distribution of the cytokine and activation of the surrounding cells , without producing toxic levels of IFN . Previous studies have implicated as limiting steps enhanceosome assembly [20] , [21] and the assembly of an interchromosomal transcriptional hub formed through interactions between Alu elements bearing NF-κB sites [20] . More recently , the infecting virus , rather than intrinsic properties of the infected cell , has been implicated in this stochasm [22] . The data presented here reveal a far more complex mechanism in which cell-to-cell variations in limiting components required to support viral replication , to detect and signal the presence of viral RNA , and to activate transcription factors all contribute to the observed stochastic expression ( Figure 7D ) . It seems likely that the key limiting factor varies between cell types , cell lines , and organisms . The earliest step in the virus induction signaling pathway is entry of virus or dsRNA into the cell . We have shown that both inducers elicit stochastic expression , but in neither case is this due to limiting inducer ( Figures S4B and 6A ) . We showed that both IFNβ-producing and nonproducing cells were infected by SeV ( Figure 2B ) . However , the IFNβ-producing cells contained significantly higher levels of the products of viral replication and transcription . Thus , it appears that there are cell-to-cell differences in the ability to support efficient viral replication , and these differences influence the probability of IFNβ gene expression . Presumably , high levels of RNA inducer in the IFN-producing cells overcome limiting amounts of RIG-I or MDA5 . However , differences in viral replication alone cannot explain the observed stochasm in IFNβ production . A previous study , using a cell line transfected with an IFNβ-GFP reporter , concluded that stochastic IFNβ expression is due entirely to heterogeneity in the infecting virus [22] . However , in that study the IFNβ-GFP cell line was preselected to minimize stochastic expression of the reporter . In addition , that study involved a stably transfected gene , while the present study made use of the endogenous gene . The results presented here strongly indicate that heterogeneity of both the virus and host cells together are responsible for the stochastic expression of IFNβ . We have identified multiple limiting steps in the activation of IFNβ gene expression , ranging from initial steps in virus infection and replication , to the signaling pathway , to the activation and binding of transcriptional activator proteins to the IFNβ promoter . For example , over-expression of individual components in the RIG-I signaling pathway increases the percentage of IFN-expressing cells . The largest increase was observed with IRF7 , which lies at the endpoint of the RIG-I pathway , and also positively controls the expression of components in the RIG-I signaling pathway . Taken together , these data are consistent with a model in which the probability of expression of the IFNβ gene in individual cells depends primarily on the activation of the RIG-I signaling pathway and the presence of sufficient numbers of IRF7 molecules to activate transcription ( Figure 7D ) . This conclusion is consistent with the observation that both IFNβ and IFNα are stochastically expressed in response to virus infection ( Figure 1A and 1C ) . The expression of both genes requires activation of the RIG-I pathway and active IRF7 [50] . We find that limiting amounts of other RIG-I pathway components also contribute to stochastic expression of the IFNβ gene , as we observed higher levels of RIG-I/Trim25 and MDA5 mRNA and protein levels in the IFNβ-producing cells than in the nonproducers ( Figure 3 ) . In addition , over-expression of RIG-I and Trim25 together leads to a dramatic increase in the percentage of cells that express IFNβ ( Figure 4C and 4D ) . Similar results were obtained with high levels of expression of the RIG-I signaling components MAVS and TBK1 and the transcription factors IRF3 and IRF7 ( Figures 4A , 4B , and 5 ) . Thus , it appears that many , if not all , of the components in the RIG-I signaling pathway , from the sensors of viral RNA to the essential transcription factors , can be limiting components in the virus induction pathway . The largest increase in the percentage of IFN-producing cells was observed when IRF7 was over-expressed . IRF7 is the master regulator of type I IFN gene expression [25] , and is present at low levels in all cell types except plasmacytoid dendritic cells , where it is constitutively abundant [26] , [40] . Our over-expression experiments show that high levels of IRF7 promote the transcription of type I IFN genes ( Figure S7D ) , and essentially eliminate the stochastic expression of both the IFNβ and α genes ( Figures 5 and S7 ) . In a previous study in human cells , both NF-κB and IRF7 over-expression was shown to partially suppress stochastic IFNβ expression [20] . Our results are consistent with this observation . However , there are two differences . First , based , at least in part , on the lack of requirement of NF-κB in murine cells , we observed a relatively small effect of increasing NF-κB expression . Second , we saw a greater effect of IRF7 expression in murine cells than was observed in human cells . Over-expression of IRF7 in L929 cells almost completely eliminated stochastic expression of both IFNβ and α genes , while in human HeLa cells high levels of IRF7 increase the percentage of IFNβ-producing cells to almost 55% [20] . Deleting the IRF7 translational repressors , 4E-BPs , also increased the IFNβ-expressing MEFs by 4-fold ( Figure 5C and 5D ) . We also showed that the RIG-I signaling pathway , and in particular RIG-I and Trim25 , are up-regulated in IRF7 over-expressing cells ( Figure S8 ) . We conclude that limiting amounts of active IRF7 appear to be overcome by two mechanisms: positive auto-regulation of IRF7 expression , and IRF7-dependent up-regulation of the RIG-I signaling pathway . We note that in addition to IFNβ , several other virus-inducible genes , including TNFα , IL-6 , CCL4 , and CCL5 , are highly expressed in the IFNβ-producing cells compared to nonproducers , suggesting that many , if not all , of the virus-inducible genes are stochastically expressed . The common feature of the activation of all of these genes is that they all require the RIG-I signaling pathway [28]–[31] . Thus , we conclude that stochastic gene expression is primarily due to limiting components in the signaling pathway but not gene-to-gene variation in the mechanism of gene activation . We showed that although the IFNβ gene is stochastically expressed upon virus infection , the antiviral ISGs , e . g . , ISG15 , were equally induced in all cells ( Figure 7C ) . However , we note that RIG-I , Trim25 , and MDA5 , which are also antiviral ISGs , are highly expressed in IFNβ-producing cells compared to nonproducing cells ( Figure 3C and 3D ) . We believe that the differences we observed here reflect naturally occurring cell-to-cell variability in the levels of expression of these genes prior to virus infection , and that this variability is the primary source of stochastic IFNβ gene expression . However , at later times after virus infection , we expect that the differences in the mRNA or protein levels of these genes between the YFP-positive and YFP-negative populations will be much smaller compared to those at earlier stages ( 8 h . p . i . ) . As shown in Figure S11A and S11B , our qPCR data and Western blot data support this expectation . The IFNβ gene is also stochastically expressed in IFNAR-deficient MEFs , which suggests that the IFNAR levels or an IFNβ feedback loop are not major factors responsible for stochastic IFNβ gene expression ( Figure S11C ) . We further measured the distributions of six components in the RIG-I signaling pathway . As shown in Figures 7A , 7B , and S10 , all six proteins were log-normally distributed across the cell population , an observation that is consistent with data on other proteins [51] , [52] . Thus , naturally occurring differences in the protein levels and activities of individual signaling pathway components and transcription factors account for stochastic IFNβ expression induced by both poly I∶C induction and virus infection . Previous studies have shown that naturally occurring differences in the levels of proteins in the apoptotic signaling pathway are the primary reasons for cell-to-cell variability in the probability of cell death [52] . Thus , the results presented here not only reveal the complexity of the regulatory mechanisms controlling stochastic IFNβ gene expression , but also suggest a general mechanism used in different biological processes to establish and control stochastic gene expression . A remarkable feature of stochastic expression is that it appears to be an intrinsic property of different clonal populations of cells . For example , if a particular cell line displays a certain percentage of activated cells , that percentage differs from other cell lines , and is retained when the cells are recloned [14] . Thus , the extent of stochasm appears to be a genetic and epigenetic feature of clonal cell populations .
All cell lines , including L929 , RAW 264 . 7 , MG63 , and 293T , were from the American Type Culture Center; primary MEFs were isolated using standard protocols from IFNβ/YFP mice [24] . Primary human foreskin fibroblast cells were purchased from PromoCell . All cells were cultured in DMEM ( Gibco ) supplemented with 10% FBS ( Gibco ) in a 5% CO2 incubator . Cycloheximide was purchased from Sigma-Aldrich . Human and mouse recombinant IFN proteins were purchased from PBL Interferonsource . Brefeldin A solution was purchased from eBioscience . Poly I∶C was purchased from InvivoGen . Cy5-labeled poly I∶C was generated using Label IT Nucleic Acid Labeling Kit ( Mirus ) . The different expression constructs were generated by cloning the coding sequences of each gene by PCR and inserting them into the vector pt-REX-DEST30 , which has the tetracycline-inducible promoter ( Invitrogen ) . Concentrated SeV stock ( Cantell strain , Charles River Lab ) was added to cultured cells at a concentration of 200 HAU/ml and incubated for the times indicated . Poly I∶C transfection was carried out using either lipofectamine2000 ( Invitrogen ) or electroporation using Amaxa MEF2 Nucleofector Kit ( Lonza ) . Total RNA was extracted with Trizol reagent ( Invitrogen ) . Real-time quantitative reverse transcription PCR ( qRT-PCR ) was conducted according to standard protocols . Antibody against YFP was from Chemicon ( Millipore ) or Abcam . RIG-I , MAVS , and GAPDH antibodies were from Cell Signaling . Antibodies against p65 , HDAC1 , and Trim25 were from Santa Cruz Biotechnology . MDA5 and TBK1 antibodies were from Abcam and Imgenex , respectively . IFNβ antibody used for FACS was from Millipore . SeV antibodies were kindly provided by Dr . Atsushi Kato ( National Institute of Infectious Diseases , Japan ) . Nuclear/cytosol fractionation was performed using Nuclear/Cytosol Fractionation Kit ( BioVision ) . Western blots were carried out using standard protocols . Antisense RNA probes recognizing mouse IFNβ or β-actin were synthesized using T7 or SP6 polymerase and digoxigenin-labeled nucleotides ( Roche Applied Science ) . Cells were cultured on poly-D-lysine-coated 24-well plates ( Fisher ) and either mock- or virus-infected for the times indicated . Cells were then washed twice with PBS and fixed with 4% paraformaldehyde . Hybridization , washes , and staining were carried out as previously described [53] . MEF cells were fixed with IC Fixation Buffer and permeabilized with Permeabilization Buffer ( both from eBioscience ) . After incubation with appropriate antibodies , flow cytometry was done with a FACSCalibur , and data were analyzed with CellQuest software ( both from Becton Dickinson ) . Total RNA from untreated and tetracycline-induced L929-IRF7 cells were prepared using Trizol reagent ( Invitrogen ) followed by purification using MEGAclear ( Ambion ) . Biotinylated RNA probes were synthesized by two rounds of amplification using the MessageAmp II aRNA Amplification kit ( Ambion ) . The probes were hybridized with Affymetrix Mouse Genome 430A_2 . 0 array chips . Affymetrix DAT files were processed using the Affymetrix Gene Chip Operating System to create CEL files . Normalized expression values were analyzed with the Bioconductor Limma package , an approach for implementing empirical Bayes linear modeling [42] . For all comparison tests , genes with an absolute fold change in transcript level exceeding 1 . 5 and p<0 . 05 were selected for further analyses . The likelihood of overrepresentation of KEGG signaling pathways in the up- or down-regulated gene list relative to a background of all array genes was calculated by Fisher's exact test for statistical analysis .
|
Eukaryotic cells can respond to extracellular signals by triggering the activation of specific genes . Viral infection of mammalian cells , for example , induces a high level of expression of type I interferons ( IFNα and β ) , proteins required for antiviral immunity that protects cells from the infection . Previous studies have shown that the expression of the IFNβ gene is stochastic , and under optimal conditions only a fraction of the infected cells express the IFNβ gene . At present neither the mechanisms nor functions of this interesting phenomenon are well understood . We have addressed this question by analyzing IFN-expressing and non-expressing mouse cells that were infected with the highly transmissible Sendai virus . We show that stochastic IFNβ gene expression is a consequence of cell-to-cell differences in limiting levels and/or activities of virus components at every level of the virus induction process , from viral replication to expression . These differences include the sensing of viral RNA by host factors , the activation of the signaling pathway , and the levels of activated transcription factors . Our findings reveal the complexity of the regulatory mechanisms controlling stochastic IFNβ gene expression . We propose that the stochastic expression of IFN allows for an even distribution of IFN , thus avoiding over-expression of IFN in infected cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology",
"microbiology",
"molecular",
"genetics",
"signaling",
"in",
"cellular",
"processes",
"biology",
"molecular",
"biology",
"immune",
"system",
"systems",
"biology",
"signal",
"transduction",
"immunity",
"virology",
"genetics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] |
2012
|
Stochastic Expression of the Interferon-β Gene
|
Bacterial metabolism has been studied primarily in liquid cultures , and exploration of other natural growth conditions may reveal new aspects of bacterial biology . Here , we investigate metabolic changes occurring when Escherichia coli grows as surface-attached biofilms , a common but still poorly characterized bacterial lifestyle . We show that E . coli adapts to hypoxic conditions prevailing within biofilms by reducing the amino acid threonine into 1-propanol , an important industrial commodity not known to be naturally produced by Enterobacteriaceae . We demonstrate that threonine degradation corresponds to a fermentation process maintaining cellular redox balance , which confers a strong fitness advantage during anaerobic and biofilm growth but not in aerobic conditions . Whereas our study identifies a fermentation pathway known in Clostridia but previously undocumented in Enterobacteriaceae , it also provides novel insight into how growth in anaerobic biofilm microenvironments can trigger adaptive metabolic pathways edging out competition with in mixed bacterial communities .
Bacteria rapidly adapt to changes in available resources and environmental fluctuations due to their remarkable ability to finely tune their physiology and metabolism [1 , 2] . Although bacterial adaptations have been primarily studied in liquid cultures composed of free-swimming planktonic cells , exploration of a broader range of growth conditions can reveal new aspects of bacterial biology [3] . One of the most common bacterial lifestyles corresponds to surface-attached communities called biofilms , in which high cell density , reduced diffusion and physico-chemical heterogeneity are associated with extensive physiological changes [4 , 5] . Biofilm bacteria display different properties compared to planktonic bacteria , and biofilms have long been considered potential reservoirs of unknown functions , contributing to their ability to thrive on surfaces present in natural and anthropic environments [6] . It was , for example , hypothesized early on that the study of physiological adjustments occurring during biofilm formation would provide insight into potentially uncharted metabolic pathways . However , due to difficulties associated with metabolic profiling in this complex environment , metabolic changes during biofilm growth are still poorly characterized . In the present study , we filtered through the wealth of molecules produced by mature Escherichia coli biofilms by restricting our analysis to comparison of volatile metabolites emitted from biofilm and planktonic cultures . We demonstrate that hypoxic conditions prevailing within biofilms induce high production of 1-propanol via a threonine fermentation pathway previously undocumented in Enterobacteriaceae , which confers a strong competitive advantage over bacteria unable to express this pathway . This study therefore demonstrates that investigation of metabolic changes associated with biofilm growth provides novel insights into the extent of bacterial metabolic potential and of bacterial adaptation to local microenvironments .
We combined headspace solid-phase microextraction ( HS-SPME ) with gas chromatography mass spectrometry ( GC-MS ) to analyze volatile compounds emitted by E . coli biofilms formed in microfermenters . Under these conditions , bacteria growing at the periphery of biofilms are exposed to oxygen present in the medium , while inner biofilm bacteria are exposed to a range of microaerobic to anaerobic conditions . We then compared volatile compounds detected in biofilms to those produced by planktonic bacteria grown aerobically ( S1 Fig ) . While these analyses revealed several differences , the most striking HS-SPME/GC-MS signal present in biofilm but absent in planktonic conditions corresponded to the volatile compound 1-propanol ( Fig 1A ) . Although low level of 1-propanol is produced by some Clostridium strains [7] , E . coli is not known to naturally produce this molecule , we therefore hypothesized the existence of a new E . coli metabolic pathway activated under conditions created by biofilm growth . Lack of 1-propanol production in aerobically grown planktonic cultures suggested that low oxygen conditions prevailing in biofilms might play a key role in E . coli 1-propanol production . Consistently , we observed 1-propanol signals in planktonic cultures grown in LB medium under fully anaerobic and microaerobic conditions , provided that the oxygen concentration did not exceed 0 . 7% ( Fig 1B ) . Conversely , when biofilms were formed in highly aerobic biofilm microfermenters , we did not detect 1-propanol ( Fig 1C ) . These results indicated that hypoxic conditions prevailing inside biofilms lead to 1-propanol production . Consistently , we showed that a mutation in the fnr gene encoding the primary transcriptional regulator mediating transition from aerobic to anaerobic growth [8] strongly reduced 1-propanol production ( S2 Fig ) . Engineering approaches used to induce 1-propanol production in E . coli often rely on the expression of heterologous alcohol dehydrogenases from various microbial sources [7 , 9] . We therefore hypothesized that at least one of the endogenous alcohol dehydrogenases of E . coli could contribute to 1-propanol production in biofilm . We tested the role of E . coli FucO , EutG , AdhE , YqhD and GlpQ alcohol dehydrogenases and determined that a mutation in the gene adhE , encoding a polymeric enzyme involved in ethanol production in E . coli and induced in anaerobic conditions [10] [11] , resulted in a marked reduction in 1-propanol production ( Fig 2A and S3 Fig ) . Lack of 1-propanol production observed in E . coli ΔadhE mutant could be restored in biofilm formed by E . coli ΔadhE complemented by pCANadhE ( S4 Fig ) . We next aimed to identify metabolic pathway constituents and necessary carbon sources for 1-propanol production . We observed that no 1-propanol could be detected in biofilms formed in LB supplemented with glucose ( Fig 2A ) , indicating that 1-propanol production is subjected to catabolic repression , nor in biofilm formed in M9 glycerol minimal media ( Fig 2B ) . However , supplementation of this medium with amino acids showed that L-threonine ( threonine ) , but not valine , serine or glycine , triggered 1-propanol production ( Fig 2B ) , as confirmed by NMR analysis ( S5 Fig ) . Consistently , growth in LB medium containing 0 . 4% threonine led to a highly significant AdhE-dependent increase in 1-propanol production ( Fig 2C ) . Finally , we monitored the steady-state level of incorporation of exogenous 13C-labeled threonine into 1-propanol in biofilm formed in M9 glycerol minimal medium . We observed strong isotope incorporation into 1-propanol , while we did not detect any significant isotope dilution by other unlabeled carbon sources ( Fig 3 ) . This demonstrated that all 1-propanol produced under these conditions originated from threonine degradation , which confirmed that threonine is the precursor of the 1-propanol pathway . Our results showed that 1-propanol can be directly produced in non-genetically-modified E . coli in a natural hypoxic microenvironment generated during biofilm growth . We observed that extension of biofilm cultures from 16 h to 48 h in biofilm microfermenters fueled with glycerol minimal medium supplemented with 0 . 4% threonine led to an increasingly strong 1-propanol signal ( S6A Fig ) . This suggested that increasing biofilm biomass correlated with increased 1-propanol production . Moreover , the 1-propanol signal produced by 24 h biofilm strongly increased when using amino-acid-rich media such as TB , a medium containing 4 times more yeast extract than LB ( S6B Fig ) . These results suggest that 1-propanol production could increase after prolonged biofilm growth in threonine-rich medium . Consistently , quantification of the amount of 1-propanol accumulated in the effluent of 48 h biofilm grown in TB or in TB supplemented with 0 . 4% threonine increased from 1 . 25 ± 0 . 15 g/L to up to 4 . 5 ± 0 . 34 g/L . These results therefore indicate that the 1-propanol yield could be optimized using amino-acid-rich unrefined media and E . coli biofilms as a production platform . In E . coli , threonine has been shown to degrade into the end products acetyl-CoA , glycine , propionate , L-isoleucine and methylglyoxal , but not 1-propanol ( Fig 4A ) . However , inactivation of genes involved in the first step of the known E . coli threonine degradation pathways , ltaE , yiaY , kbL , ilvA and tdcB , showed that only strains harboring a mutation in the threonine dehydratase gene tdcB impaired 1-propanol production in biofilm ( Fig 4B and S1 Table ) . The lack of 1-propanol production observed in E . coli ΔtdcB mutant could be restored in biofilm formed by E . coli ΔtdcB complemented by pCANtdcB ( S4 Fig ) . tdcB is part of the tdc operon , which is negatively regulated by catabolic repression and is involved in anaerobic uptake and degradation of threonine to propionate ( Fig 4A ) [12] . We tested the contribution of the other genes coding for enzymes involved in aerobic ( pflB , ptA and ackA ) and anaerobic ( tdcC , tdcE , ptA and tdcD ) propionate production , and showed that their inactivation did not affect 1-propanol production , except for tdcC ( encoding a threonine uptake transporter ) and tdcE ( encoding a 2-ketobutyrate formate-lyase ) ( S1 Table and S7A and S7B Fig ) . QRT-PCR gene expression analysis confirmed the strong induction of tdcBDE genes in planktonic anaerobic conditions ( 1 , 500-fold ) and biofilms ( 300-fold ) grown in LB medium ( S8A and S8B Fig ) . AdhE catalyzes successive reduction of acetyl-CoA to acetaldehyde and the latter compound to ethanol [11] . We hypothesized that the promiscuous AdhE enzyme , which is also induced in biofilm conditions ( S8C Fig ) , could carry out successive reduction of propionyl-CoA into propionaldehyde ( coenzyme-A-dependent aldehyde dehydrogenase activity of AdhE ) , and then , reduction of propionaldehyde into 1-propanol ( alcohol dehydrogenase activity of AdhE ) ( Fig 4C ) . Consistent with the existence , in E . coli , of a metabolic pathway branching out from the propionate pathway at the level of propionyl-CoA , we observed AdhE-dependent , increased production of 1-propanol in biofilms upon supplementation of LB medium with propionaldehyde ( S9A Fig ) . We also observed concomitant increased production of ethanol in a tdcB mutant ( S9B Fig ) , suggesting that blocking conversion of threonine into 1-propanol redirects AdhE metabolic activity towards ethanol synthesis . We next examined conservation of this newly uncovered threonine degradation pathway among bacterial taxa . We determined that co-occurrence of homologs of adhE , tdcB and tdcE genes required for threonine degradation into 1-propanol can be identified in many bacteria , with the strongest homology found in Enterobacteriaceae ( S10 Fig ) . Consistently , all tested E . coli isolates naturally produced 1-propanol in biofilms , while several other Enterobacteriaceae species , including Shigella flexneri , Salmonella enterica sv . Enteritidis and Citrobacter rodentium , also produce 1-propanol in anaerobic , but not in aerobic planktonic cultures ( Fig 5 ) . Since E . coli K-12 cannot use 1-propanol as a carbon source , and exposure to 1-propanol did not display a detectable phenotype ( S11 Fig ) , what could be the function of this threonine degradation into 1-propanol ? Considering that reduction of propionyl-CoA into 1-propanol involves two successive steps of re-oxidation of reduced nicotinamide adenine dinucleotide ( NADH ) into NAD+ ( Fig 4C ) , we hypothesized that degradation of threonine into 1-propanol recycles NADH into NAD+ , a key co-factor playing a major role in central metabolism [11] . Indeed , analysis of the NADH/NAD+ ratio under biofilm and planktonic anaerobic growth conditions showed that bacteria exhibit an increased NADH/NAD+ ratio in ΔtdcB and ΔadhE mutants , but not in ΔtdcD mutants ( Fig 4D and S12A , S12B and S12C Fig ) . Hence , while production of 1-propanol from threonine was reported in some Clostridia strains , it also corresponds to a previously undescribed native fermentation pathway in E . coli , contributing to intracellular redox balance in conditions of energy starvation in the absence of oxygen . To investigate the biological consequences of threonine fermentation in E . coli , we performed competition experiments between E . coli WT and a ΔtdcB ( no fermentation of threonine into 1-propanol ) or a ΔtdcD mutant ( threonine fermentation into 1-propanol ) , either in biofilm or planktonic anaerobic conditions . Whereas the tested strains did not display any growth defect in monoculture ( Fig 6A and 6B ) , we observed a 90% fitness reduction in the ΔtdcB mutant in competition experiments against WT when grown either in biofilm or planktonic anaerobic conditions ( Fig 6A , 6B and 6C ) , whereas the tdcD mutant displayed no growth nor fitness defect ( Fig 6B ) . Similarly , a ΔtdcB mutant display fitness reduction compared to WT , in competition experiments against Klebsiella pneumoniae ( Fig 6D ) , a strain that does not ferment threonine into propanol . The fitness defect of a ΔtdcB mutant correlates with a consistent growth lag in planktonic anaerobic conditions compared to a wild-type or unaffected ΔtdcD mutant , ( Fig 6E and S13 Fig ) . In contrast , a tdcB mutant did not display any growth lag ( Fig 6F ) or fitness cost when competition experiments were performed in aerobic planktonic culture conditions ( Fig 6G ) , demonstrating the particular relevance of this novel pathway during anaerobic and biofilm growth . These results demonstrate that the threonine-to-propanol fermentation pathway contributes to provide a competitive advantage in mixed anaerobic communities .
We show that native production of 1-propanol is not restricted to a few anaerobic bacteria , but naturally occurs from threonine degradation under hypoxia in E . coli and other Enterobacteriaceae . The link between 1-propanol production and threonine catabolism was previously reported in Clostridium sp . strain 17cr1 , which produces low amounts of 1-propanol ( less than 70 mg/L ) by an uncharacterized pathway [13] . Moreover , engineered E . coli strains can produce 1-propanol upon reduction of 2-keto-butyrate formed by the aerobic threonine degradation pathway [14] . In E . coli , we show here that anaerobic reduction of propionyl-CoA into propanal and 1-propanol by alcohol/aldehyde dehydrogenase AdhE constitutes a native alternative to anaerobic degradation of threonine into propionate and ATP synthesis by enzymes encoded by the tdc operon [12] . Although 1-propanol can be produced in anaerobic planktonic cultures , this metabolic capacity might have been overlooked due to the low biomass reached in these conditions . In contrast , we hypothesize that hypoxia spontaneously developing in biofilms enables a large bacterial biomass to be exposed to optimal conditions , inducing high production of 1-propanol . 1-propanol is an important industrial solvent and a major component of resins , the chemical synthesis of which requires a laborious two-step process involving catalytic hydroformylation of ethylene to produce propanal , and consecutive hydrogenation of propanal into 1-propanol [15–17] . Hence , improvement in renewable biological production of 1-propanol in metabolically engineered E . coli strains carrying heterologous genes of varying microbial origin recently gained significant attention [18 , 19] . We determined that the yield of 1-propanol spontaneously produced in continuous flow biofilms in amino-acid-rich media could reach up to 5 g/L in the biofilm effluent , a yield close to levels obtained using engineered strains grown in discontinuous fed-batch cultures ( <10 g/l ) [7] . Achieving industrial productivity using bacterial biomass immobilization in biofilm reactors still presents many important obstacles and challenges [20] . However , our results suggest that bioproduction of 1-propanol from a native , and therefore robust , metabolic pathway induced in E . coli biofilms could alleviate the need for establishing synthetic pathways in genetically modified organisms , and might constitute an alternative to current 1-propanol chemical synthesis . Fermentation is a central metabolic process that has been thoroughly investigated in Enterobacteriaceae in a variety of planktonic culture conditions , leading to production of lactate , ethanol , acetate , formate , citrate , succinate , hydrogen and carbon dioxide . However , our study reveals that , in addition to these classical fermentation products , E . coli and other Enterobacteriaceae can also produce 1-propanol upon reduction of threonine and reoxydation of NADH into NAD+ , a staple of the fermentation process . The ability to ferment threonine as well as aromatic and branched-chain amino acids via the Stickland reaction is particularly used by anaerobic bacteria such as Clostridia [21–23] . This reaction is characterized by oxidation of one amino acid coupled with reduction of another amino acid [24] . However , we did not observe any stimulation of 1-propanol production upon supplementation with various amino acids , suggesting that threonine-to-propanol fermentation does not correspond to a bona fide Stickland reaction [25] . We show that reduction of threonine contributes to cellular redox balance by restoring the intracellular oxydized NAD+ pool , which may play an important role in E . coli’s ability to cope with anaerobic environments . Consistently , lack of threonine reduction into 1-propanol in a tdcB mutant leads to a remarkable 90% fitness loss in competition with the wild type strain . This decreased fitness could be attributed to a marked growth lag , enabling depletion of limited nutrient resources by the competing wild type strain . In the context of biofilm formation , maintaining redox balance is likely to be an essential metabolic process . Consistently , in absence of threonine fermentation , E . coli biofilm bacteria can use another fermentation pathway to recycle consumed NADH into NAD+ , as shown by increased ethanol fermentation observed in a tdcB mutant ( S9B Fig ) . However , these pathways generally rely on the availability of glucose or other oxidized sugars , which , unless produced by costly gluconeogenesis , may not be generally abundant in E . coli nutritional environment . By contrast , the amino-acid fermentation pathway described in this study could confer Enterobacteriaceae the ability to maintain redox balance and edge out competition with other bacteria , using amino-acids produced by cell lysis and proteolysis inside biofilms . Another of such amino-acid rich environment could correspond to biofilm-like anaerobic gut environments , in which glycosylated mucins abundantly secreted in epithelial mucus contain up to 40% serine and threonine [26 , 27] . However , in vivo mouse colonization competition experiments between E . coli WT and tdcB mutants did not reveal any significant colonization defect ( S14 Fig ) , which might simply reflect the fact that the competitive fitness advantage provided by the propanol pathway cannot be relevantly tested in largely herbivorous mice . Identification of a widespread metabolic response to biofilm and anaerobic conditions expands the range of known E . coli metabolites , opening perspectives of biofilm-based approaches for harnessing bacterial metabolic potential . Our study also further supports the notion that mining of the biofilm mode could provide insight into new aspects of bacterial physiological adaptations to local microenvironments .
Bacterial strains used in this study are listed in S2 Table . E . coli mutants listed in S1 Table are from the Keio Collection [28] and each mutation was introduced into the E . coli TG1 strain by P1 vir phage transduction . Each mutant was confirmed by PCR . All experiments were performed in: lysogeny broth ( LB ) containing as amino acid sources 1% peptone and 0 . 5% yeast extract; Terrific broth ( TB ) containing 1 . 2% peptone , 2 . 4% yeast extract; or M9 glycerol 0 . 4% minimal medium containing no amino acid source . These media were supplemented with kanamycin ( 50 μg/ml ) when required and incubated at 37°C . When needed , 0 . 4% ( wt/vol ) L-threonine ( indicated as threonine throughout the text ) , glycine , serine or valine was added to the cultures . All media and chemicals were purchased from Sigma-Aldrich . Growth under anaerobic and microaerobic conditions was performed in a C400M Ruskinn anaerobic-microaerophilic station on multi-position magnetic stirrers XT35 . 1 ( Roth Sochiel ) at 37°C . Continuous-flow biofilm microfermenters containing a removable glass spatula were used as described in [6] ( see also https://research . pasteur . fr/en/tool/biofilm-microfermenters/ ) in one of the following methods: Biofilm microfermenters were inoculated by placing the spatula in a culture solution adjusted to OD600 = 1 ( containing 5 . 108 bacteria/ml ) for 5 min . The spatula was then reintroduced into the microfermenter . Flow rate was then adjusted ( 60 ml/h ) so that total time for renewal of microfermenter medium was lower than bacterial generation time , thus minimizing planktonic growth by constant dilution of non-biofilm bacteria . Total RNA was extracted from three independent samples using the Qiagen RNeasy mini-kit . DNase treatment on 3 mg of RNA was carried out twice with the Ambion Turbo DNA-free kit . All samples were checked for residual genomic DNA contamination with the TM1 and TM2 primer pair ( S3 Table ) , and were considered DNA-free if no amplification was detected at <38 cycles . The RNA concentration was measured with a NanoDrop and RNA quality was checked by gel electrophoresis . cDNA synthesis was carried out with 2 μg of RNA in a volume of 50 μl using the Bio-Rad iScript cDNA synthesis kit . Primers for quantitative real-time PCR were designed using the Primer3 online tool ( http://simgene . com/Primer3 ) and are listed in S3 Table . Amplicon sizes were confirmed by gel electrophoresis . cDNA levels were analyzed by EvaGreen detection in a Bio-Rad CFX96-1000 light cycler using the Bio-Rad SoFast EvaGreen Supermix ( 20 μl final volume ) , with 200 nM of each primer . Melting curves were checked to confirm that a single PCR product had been amplified . All quantitative real-time PCR reactions were carried out in quadruplicate for each sample in 96-well plates with simultaneous no-template controls . Relative quantification of gene expression levels was determined with the Delta Delta CT method [29] using 16S ( rssH ) , ihfb , opgG and hcaT genes as reference genes . E . coli TG1 was grown in M9 glycerol 0 . 4% minimal medium supplemented with 0 . 2% L-threonine . The culture sample was collected after 48 h , centrifuged and the supernatant was frozen at -20°C prior to NMR analysis . 1D and 2D NMR spectra were recorded on a Bruker Avance III 800 MHz instrument ( Bruker , Bremen , Germany ) , equipped with a 5 mm QPCI ( 1H , 13C , 31P and 15N ) cryoprobe . Supernatants ( 120 μl ) were mixed with 40 μl of 1 mM 2- ( trimethylsilyl ) propionic-2 , 2 , 3 , 3-d4 acid ( TSP-d4 ) solution in D2O as an internal intensity and chemical shift standard , without further sample pretreatment . Data were acquired and processed using TOPSPIN 3 . 0 software . E . coli biofilm was grown on M9 glycerol 0 . 4% minimal medium , supplemented with a mixture ( 1:1 ) of unlabeled L-threonine and [U-13C]L-threonine , in which the total concentration of L-threonine was 2 g/L . After 48 h of E . coli TG1 growth in the microfermenter at 37°C , the flux was stopped , and 15 h later , the biomass of the biofilm was recovered and centrifuged and the supernatant frozen until NMR analysis; Threonine 13C-enrichment was measured by NMR to be 44 . 5% ± 0 . 2% . The occurrence of a 13C atom resulted in splitting of 1H resonance due to 1H-13C coupling into 1H-12C ( central peak ) and 1H-13C ( “satellite” peaks ) signals . The percentage of 13C incorporated into the carbon position was measured by the area of 1H-13C signals relative to total resonance area . Specific 13C-enrichments on C2 ( 41 . 4% ± 0 . 6% ) and C3 ( 42 . 4% ± 0 . 2% ) of 1-propanol indicated the absence of significant isotopic dilution of 13C -labeled threonine when converted into 1-propanol . Insert: carbon positions in 1-propanol . Extraction of NADH and NAD+ was carried out according to the method described in [30] . Pellet samples of anaerobic planktonic culture or biomass of the biofilm were centrifuged at 16 , 000 x g for 1 min . Supernatant was removed and pellets were resuspended in 300 μl of 0 . 2M NaOH ( for NADH extraction ) or 0 . 2M HCl ( for NAD+ extraction ) . These extracts were incubated for 10 min at 50°C and then for 10 min on ice . While vortexing , 300 μl of 0 . 1M HCl ( for NADH ) or 0 . 1M NaOH ( for NAD+ ) was added dropwise to neutralize the solutions . Samples were then centrifuged for 5 min at 16 , 000 x g . Supernatants were transferred to fresh tubes and stored at -80°C until quantification . Relative or absolute NADH or NAD+ levels were quantified using an enzyme cycling assay adapted for measurement in a microtiter plate [31] . A master reagent mix was prepared with 1 x bicine buffer ( 1 . 0 M pH 8 ) , 3 x water , 1 x 40 mM EDTA , 1 x 100% ethanol , 1 x 4 . 2 mM thiazolyl blue and 2 x 16 . 6 mM phenazine ethosulfate . The reagent mix was warmed to 30°C , and then 90 μl aliquots were dispensed into individual wells of a 96-well microtiter plate; 5 μl of standard or sample were added to each well and the plate was incubated for approximately 10 min at 30°C . Then , the cycling reaction was started by the addition of 5 μl of alcohol dehydrogenase ( Sigma no . A-3263 ) prepared at 347 units/ml in 0 . 1M bicine ( pH8 . 0 ) . The microtiter plate was incubated at 30°C , contents were mixed by brief shaking and absorbance was measured every 60 s using a TECAN Infinite M200 PRO at 570 nm , i . e . the spectral peak of thiazolyl blue that increases upon reduction . Slopes arising from plots of absorbance at 570 nm over time were generated for NADH and NAD+ standards , as well as for all samples . Standard curves were used to calculate absolute concentrations in μM , and values were normalized to the optical density of the original cell culture sample . Volatile compounds emitted by bacterial planktonic or biofilm cultures were determined using an analytical approach coupling headspace solid phase microextraction ( HS-SPME ) with gas chromatography and mass spectrometry ( GC-MS ) [33] [34] . Preparation of biofilm extracts for HS-SPME/GC-MS analysis was as follows . After 24 h of culture at 37°C , biofilm biomass that had formed on the spatula was scraped off , put in a 10 ml headspace vial and frozen until HS-SPME/GC-MS analysis ( see S1 Fig ) . 1-Propanol was purchased from Sigma-Aldrich ( Saint Quentin Fallavier , France ) . Ultra-pure water was produced using a Direct-Q UV 3 system ( 18 . 2 MΩ/cm ) from Millipore ( Molsheim , France ) ; 75 μm carboxen-polydimethylsiloxane ( CAR-PDMS ) fiber was from Supelco ( Sigma-Aldrich , Saint Quentin Fallavier , France ) was used for SPME . The fiber used was conditioned prior to performing analyses by inserting them into the GC injector at 280°C for 10 min . For each HS-SPME analysis , equivalent planktonic or biofilm bacterial biomass resuspended into 1 ml of planktonic or biofilm medium was introduced into 10 mL SPME vials . The fully automated HS-SPME procedure was as follows . First , the vial was equilibrated at 60°C for 6 min; then , the SPME fiber was placed into the head-space of the sample vial for extraction and maintained at 60°C for 30 min . At the end of extraction , the SPME fiber was introduced directly into the GC injector ( desorption ) for 10 min at 280°C in split mode ( ratio1:2 ) . GC-MS analyses were performed on an Agilent 7890A gas chromatograph coupled with an Agilent detector 5975C inert XL MSD mass spectrometer ( Agilent Technologies , Les Ulis , France ) . The device is equipped with an MPS autosampler from Gerstel ( RIC , Saint-Priest , France ) that enabled fully automated HS-SPME analyses . The column used was a non-polar ( methyl 95%-phenyl 5% ) fused silica capillary column CP-SIL 8CB-MS ( 30 m x 0 . 25 mm with 1 μm film thickness ) obtained from Agilent Technologies ( Les Ulis , France ) . Helium was used as carrier gas in constant flow mode at 1 mL/min . The injector temperature was 280°C; injection mode was in split mode with a split ratio of 1:2 . The temperature program was 40°C , held for 3 min and then raised to 60°C at 2°C/min , increased to 300°C at 20°C/min and held for 3 min ( run 28 min ) . The transfer line temperature to the MS detector was set at 280°C . A mass spectrometer was used with the positive electronic ionization ( EI ) source ( 70 eV ) heated to 230°C and the MS quad at 150°C . Acquisition was simultaneously performed in scan and SIM ( single ion monitoring ) modes . Scan acquisition was made from m/z 20 to 250 . For SIM acquisition , the m/z fragments selected as characteristic fragment ions of 1-propanol were 31 ( CH2 = OH+ ) and 59 ( CH3-CH2-CH2O+ ) . Identification of volatile organic compounds was performed by matching their recorded mass spectra with standard mass spectra from the National Institute of Standards and Technology ( NIST , ver 2 . 0f , rev . 2010 ) . The identity of each volatile compound was also confirmed by comparing their retention time and mass spectra with those of pure standard compounds after GC-MS analysis . The yield of 1-propanol in biofilm was determined either by resuspending 24 h biofilms formed on the internal microfermenter directly in microfermenter medium , or by resuspending 24 h biofilms for 15 or 24 h or 48 h after stopping medium flow , to allow accumulation of 1-propanol . The 1-propanol yield in anaerobic planktonic culture was tested by direct sampling of culture medium . After centrifugation , 20 ml of biofilm resuspension or planktonic culture supernatant were sent to Aromalyze ( Quetigny , France http://www . aromalyse . fr ) . All samples were homogenized by vigorous stirring prior to analysis , and an aliquot of 0 . 5 ml was transferred into a 20 ml glass vial with a magnetic cap equipped with a PTFE/silicon septum containing 7 . 5 ml of deionized water and 2 . 5 g of NaCl ( 99% , Aldrich ) . As internal standard , an aqueous solution of 1-propanol-d7 ( 98atom% d , CDN Isotopes ) at known concentration was added . All samples were analyzed by HS-SPME/GC-MS carried out at 40°C for 30 min using a carboxen-polydimethylsiloxane ( CAR-PDMS ) fiber ( Supelco ) on a CombiPAL . Fiber desorption was performed in the injector unit of the chromatograph in splitless mode . GC-MS analyses were carried out on a Shimadzu 2010 chromatograph coupled with a Shimadzu QP2010+ mass spectrometer . The capillary column used was an Rtx-624 ( Restek ) with stationary phase cyanopropylphenylated ( 6% ) dimethyl polysiloxane ( 94% ) . The carrier gas was helium . Oven programming was as follows: 30°C for 5 min , 10°C/min to 150°C , 20°C/min to 300°C . Data acquisition was done in scan mode ( m/z = 29 to 100 , electron impact , 70eV ) . A blank sample ( water replacing sample aliquot ) was analyzed after each sample to exclude cross-contamination . All samples were analyzed in duplicate; several arbitrarily chosen samples were analyzed as triplicates . In all cases , the coefficient of variation on the calculated 1-propanol content was < 1 . 5% . Quantification was done using a signal of 1-propanol-d7 ( isotope dilution analysis ) and ions 60 for 1-propanol and 67 for 1-propanol-d7 . Paired or unpaired Student t-test analyses were performed using Prism 6 . 0 for Mac OS X ( GraphPad Software , Inc . ) . Each experiment was performed at least 3 times . * p<0 . 05; ** p<0 . 01; *** p<0 . 001 .
|
Whereas Escherichia coli does not naturally produce the 1-propanol unless subjected to extensive genetic modifications , we show that this important industrial commodity is produced in hypoxic conditions inside biofilms . 1-propanol production corresponds to a native threonine fermentation pathway previously undocumented in E . coli and other Enterobacteriaceae . This widespread adaptive response contributes to maintain cellular redox balance and bacterial fitness in biofilms and other amino acid-rich hypoxic environments . This study therefore shows that mining complex lifestyles such as biofilm microenvironments provides new insight into the extent of bacterial metabolic potential and adaptive bacterial physiological responses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"bacteriology",
"biofilms",
"chemical",
"compounds",
"metabolic",
"processes",
"biological",
"cultures",
"microbiology",
"threonine",
"organic",
"compounds",
"fermentation",
"enterobacteriaceae",
"amino",
"acids",
"bacteria",
"research",
"and",
"analysis",
"methods",
"amino",
"acid",
"metabolism",
"proteins",
"chemistry",
"anaerobic",
"bacteria",
"biochemistry",
"bacterial",
"biofilms",
"organic",
"chemistry",
"hydroxyl",
"amino",
"acids",
"biology",
"and",
"life",
"sciences",
"cell",
"culturing",
"techniques",
"physical",
"sciences",
"biofilm",
"culture",
"metabolism",
"organisms"
] |
2017
|
Biofilm microenvironment induces a widespread adaptive amino-acid fermentation pathway conferring strong fitness advantage in Escherichia coli
|
Intracellular polarization , where a cell specifies a spatial axis by segregation of specific factors , is a fundamental biological process . In the early embryo of the nematode worm Caenorhabditis elegans ( C . elegans ) , polarization is often accompanied by deformations of the cortex , a highly contractile structure consisting of actin filaments cross-linked by the motor protein myosin ( actomyosin ) . It has been suggested that the eggshell surrounding the early embryo plays a role in polarization although its function is not understood . Here we develop a mathematical model which couples a reaction-diffusion model of actomyosin dynamics with a phase field model of the cell cortex to implicitly track cell shape changes in the early C . elegans embryo . We investigate the potential rigidity effect of the geometric constraint imposed by the presence and size of the eggshell on polarization dynamics . Our model suggests that the geometric constraint of the eggshell is essential for proper polarization and the size of the eggshell also affects the dynamics of polarization . Therefore , we conclude that geometric constraint on a cell might affect the dynamics of a biochemical process .
The geometry of a cell can have a profound influence on cell function and survival [1–4] . Cell geometry is generated by internal structures such as the cytoskeleton , and can also be externally imposed by interactions with neighboring cells or mechanical structures like an eggshell . Eggshells are critical for early development of the nematode worm C . elegans as it prevents multiple sperm from fertilizing a single egg , along with ensuring proper chromosome segregation during meiosis , and proper organization of membrane and cortical proteins [5–7] . Shortly after fertilization , the C . elegans embryo polarizes by asymmetrically localizing specific proteins , including actin , myosin and polarity determinants such as the Par proteins , in response to a cue from the sperm [8 , 9] . Polarization of the embryo proceeds in two distinct phases: establishment and maintenance [10] . During the establishment phase , the developmental time frame we consider in this paper , the actomyosin cortex , a thin structure below the membrane consisting primarily of polymerized actin filaments and cross-linked by the motor protein myosin , is highly dynamic and contractile , creating small invaginations on the cell surface called ruffles [9 , 11] . The cue locally relaxes the actomyosin cortex causing local loss of ruffles and initiation of cortical flow that transports the anterior Par proteins , PAR-3 , PAR-6 and aPKC , towards the anterior pole . This allows the posterior Par proteins , PAR-1 , PAR-2 and LGL , which are mutually antagonistic to the anterior Par proteins , to bind to the cleared area at the posterior pole [10 , 12 , 13] . During advection of the Par proteins and the actomyosin cap , a domain of high actomyosin density , towards the anterior pole , an invagination similar to the ruffles but much deeper , called the pseudocleavage furrow , forms and moves with the edge of the actomyosin cap and the interface between the anterior and posterior Par proteins [9 , 10 , 14] . At the end of the establishment phase , the pseudocleavage furrow along with the actomyosin cap and the Par protein interface reach the middle of the cell . The pseudocleavage furrow retracts and the segregated Par protein domains are held through the maintenance phase as the cell prepares for first division . It has been demonstrated that the pseudocleavage furrow and ruffles are not essential for proper polarization . A maternal effect mutation nop-1 ( it142 ) eliminates cortical contractility during the establishment phase which results in the absence of the pseudocleavage furrow and ruffles but does not appear to interfere with proper development despite attenuation of actomyosin asymmetry and cortical flow [15 , 16] . Although it is known that the pseudocleavage furrow and ruffles are generated as a result of contractility , and the pseudocleavage furrow appears at the boundary between high and low actomyosin concentrations , the mechanical generation of these invaginations is not fully understood . The developing C . elegans embryo is surrounded by an eggshell , a rigid body containing the protein chitin , which is created by the embryo after fertilization and persists until hatching . When the eggshell is genetically or chemically manipulated prior to the two cell stage , the cells are highly mechanically sensitive and fail to properly extrude polar bodies at the beginning of the establishment phase , and do not properly polarize or form the pseudocleavage furrow [5–7] . Removal of the eggshell at or after the two cell stage does not interfere with development [17] . Since chitin is responsible for eggshell rigidity [5] , this suggests that the eggshell may provide structural support during the asymmetric actomyosin contractions associated with polarization [5] , although this role is not well understood . In this paper , we develop a mathematical model to investigate the mechanical effect of the eggshell during establishment of polarization . This model allows us to modify or remove the eggshell completely to explore the role of the eggshell shape and rigidity on polarization . The model simulates the morphological changes of the membrane during establishment of polarization which includes movement of the pseudocleavage furrow . Reaction-diffusion equations describing the dynamics of Par proteins and actomyosin [1] are incorporated into the moving interface problem . Cell shape dynamics are tracked using the phase field method which is widely used for moving interface problems [18–20] . In the phase field method , the interface is not explicitly tracked , which avoids potential numerical difficulties associated with the classical Lagrangian method . Here the interface is the level set of an auxiliary field ϕ . ϕ takes on different constant values corresponding to inside and outside of the cell and these constant values are connected by a smooth interface . In our model , cortex is assumed to be a viscous material and the asymmetric actomyosin contractions in the cortex drives cell deformations . The interactions between membrane bending energy , surface tension , eggshell constraint and forces applied by the cortex determines the current state of the cell . Using numerical simulations we found that the eggshell rigidly supports the membrane and reinforces the polarization process . We also showed the importance of the size of the eggshell for proper polarization . Our results suggest that mechanical constraints on cells imposed by structures such as an eggshell might affect dynamics and localization of proteins .
In the previous work by Dawes et al . [1] , a model of reaction-diffusion equations was developed to capture the dynamics of Par proteins and actomyosin on a fixed domain . In their model , cortical tension is assumed to be linearly proportional to actomyosin concentration and the advective speed of protein movement is linearly proportional to the gradient of cortical tension . The solution is shown to have a moving interface between high and low concentrations of the proteins , which represents the movement of the interface between the anterior and posterior domains during the protein movement . While the model in [1] is capable of capturing the essential features of the protein dynamics with relatively simple equations , it does not include any change in cell shape and therefore is limited in providing information about the physics of the entire polarization process . Here we consider the same species and reactions , and along with protein dynamics , we include the formation and movement of the pseudocleavage furrow as well as the constraint of the eggshell . In our model , since the plasma membrane and cortex are mostly in contact and deforming together , we do not distinguish the plasma membrane and cortex but assume they are a single structure , and in the following we use membrane or cortex to refer to this structure . Following the model in [1] , we do not consider individual Par proteins but group these proteins into two modules by their localizations: the anterior and posterior Par proteins . The anterior Par protein module consists of PAR-3 , PAR-6 and aPKC , and the posterior Par protein module consists of PAR-1 , PAR-2 and LGL . Thus , the model contains the following species with the corresponding notations: Am: Cortical anterior Par protein monomers , Asd: Cortical anterior singly bound dimers , Add: Cortical anterior Par protein doubly bound dimers , P: Cortical posterior Par proteins , M: Cortical actomyosin , Ay: Cytoplasmic anterior Par protein monomers , A2y: Cytoplasmic anterior Par protein dimers , Py: Cytoplasmic posterior Par proteins My: Cytoplasmic actomyosin . We make the following key assumptions about the interactions as in [1 , 21]: The anterior Par proteins can dimerize and bind to the cortex , and the anterior and posterior Par proteins promote each other’s dissociation from the cortex via phosphorylation . The interactions between these species follow mass-action kinetics . Cytoplasmic Par proteins , Ay , A2y and Py , are assumed to be at quasi-steady state and can associate with the cortex , and proteins dissociated from the cortex return to the cytoplasmic pool . Cytoplasmic actomyosin My is assumed to be at quasi-steady state . Cytoplasmic actomyosin assembles into the cortex and this assembly is negatively regulated by the posterior Par proteins . The interactions between these species take place on the cortex , and therefore the domain of interest is the evolving cortex . A schematic diagram of the protein interactions are depicted in Fig 1C , and the protein dynamics can be described by the following equations: ∂ [ A m ] ∂ t = D p a ∇ c 2 [ A m ] - ∇ c · ( v c [ A m ] ) + k on A A y - k off A [ A m ] - 2 k d + [ A m ] 2 + 2 k d - [ A d d ] - k d + A y [ A m ] + k d - [ A s d ] - r A [ P ] [ A m ] , ( 1 ) ∂ [ A s d ] ∂ t = D p a ∇ c 2 [ A s d ] - ∇ c · ( v c [ A s d ] ) + k on A s d A 2 y - k off A [ A s d ] + k d + A y [ A m ] - k d - [ A s d ] - k on A d d [ A s d ] + k off A [ A d d ] - r A [ P ] [ A s d ] , ( 2 ) ∂ [ A d d ] ∂ t = D p a ∇ c 2 [ A d d ] - ∇ c · ( v c [ A d d ] ) + k d + [ A m ] 2 - k d - [ A d d ] - k off A [ A d d ] + k on A d d [ A s d ] - 2 r A [ P ] [ A d d ] , ( 3 ) ∂ [ P ] ∂ t = D p p ∇ c 2 [ P ] - ∇ c · ( v c [ P ] ) + k on P P y - k off P [ P ] - r P ( [ A m ] + [ A s d ] + 2 [ A d d ] ) [ P ] , ( 4 ) ∂ [ M ] ∂ t = D m ∇ c 2 [ M ] - ∇ c · ( v c [ M ] ) + k on M M y k P k P + [ P ] - k off M [ M ] . ( 5 ) In Eqs ( 1 ) – ( 5 ) , ∇ c 2 stands for the Laplace-Beltrami operator and ∇c is the gradient operator on the membrane . The anterior Par proteins , posterior Par proteins and actomyosin diffuse on the cortex at the rates Dpa , Dpp and Dm , respectively . The protein movement induced by the initial cue which relaxes the cortex at the posterior pole drives the advection of anterior and posterior Par proteins . The advective velocities for all of the proteins , denoted by vc , are assumed to be linearly proportional to the gradient of cortical tension [1] , which we assume to depend linearly on the actomyosin concentration , that is , vc = ν∇M with ν a constant . The Par proteins move along with the cortex , and therefore the aforementioned protein dynamics occurs on a deforming interface . To model the evolution of the cortex , we use the phase field model which implicitly tracks moving interfaces . The choice of phase field model over other explicit interface tracking methods , such as a Lagrangian frame , is based on the ease of coupling the protein dynamics with the deforming cortex , while avoiding numerical difficulties that may arise due to the deep invagination of the pseudocleavage furrow . Since we are interested in the dynamics at the cell cortex , we use a 2D cross section of the cell as our domain Ω ⊂ R 2 , as shown in Fig 1A . We believe that this model , albeit simplified in geometry , will provide information on the possible mechanisms underlying cell-shape changes and pseudocleavage furrow formation . The phase field function ϕ ( x ) is defined on the entire domain Ω , where the region with ϕ = 0 represents the exterior of the cell , the region with ϕ = 1 represents the interior of the cell , and the region with 0 < ϕ < 1 is the “diffuse interface” including the level set ϕ = 0 . 5 that represents the cell membrane Ωs . The width of this diffuse interface is controlled by the transition parameter ϵ , which is taken to be a very small number . Note that the choice of ϵ depends on the spatial resolution: the spatial grid needs to be fine enough to resolve the diffuse interface . The evolution of the diffuse interface is governed by kinetic equations that is defined on the whole domain , and they will be described in the remainder of this section . Given a fixed membrane profile ( i . e . , a fixed ϕ ) , one can easily simulate the protein dynamics of Eqs ( 1 ) – ( 5 ) along the cell membrane by introducing G ( ϕ ) = 18ϕ2 ( ϕ − 1 ) 2 in the equations . Note that G ( ϕ ) , the double well potential , is nonzero only around the membrane , and therefore by multiplying G with the unknown variables or reaction terms , the protein dynamics are restricted to the neighborhood of the interface . In our numerical simulations , a non-dimensionalized version of model ( 1 ) – ( 5 ) was used . The non-dimensionalized model equations , coupled with G ( ϕ ) , are displayed in Eqs . ( S1 ) - ( S5 ) of S1 Text . In our model , the cell shape , i . e . , the phase field function ϕ , is determined by the interactions between 1 ) the membrane bending energy , 2 ) membrane surface tension , 3 ) volume conservation , 4 ) eggshell constraint and 5 ) the force resulting from myosin contractility in aligned and cross-linked forms . Here we assume that that the cortex and cytoplasm are viscous with different viscosities , which is appropriate for long time scales [22–24] . The phase field function ϕ is governed by the following equation [19 , 25]: ∂ ϕ ∂ t + u · ∇ ϕ = Γ ( ϵ ∇ 2 ϕ - G ′ ( ϕ ) / ϵ + c ϵ | ∇ ϕ | ) , ( 6 ) where u is the velocity field driving the membrane-cortex deformation , Γ the relaxation coefficient and c = ∇ ⋅ ( ∇ϕ/|∇ϕ| ) the local interface curvature , which is added to stabilize the phase-field interface [19] , ϵ∇2 ϕ − G′ ( ϕ ) /ϵ is boundary free energy , which guarantees the existence of two phases connected through a smooth interface . The velocity field u is generated by the forces on the membrane , and it is also associated with the viscosity of the cortex and cytoplasm . Assuming that viscosity forces dominate advective inertial forces [23] , u is the solution of the Stokes equation: ∇ · [ η ( ϕ ) ( ∇ u + ∇ u T ) ] - ξ u + F m e m = 0 . ( 7 ) In Eq ( 7 ) , the first term describes the strain rate tensor where η ( ϕ ) is the viscosity . It is defined as η ( ϕ ) = ηm m4ϕ ( 1 − ϕ ) + ηc ϕ , where m is the non-dimensionalized variable for actomyosin concentration [M] , 4ϕ ( 1 − ϕ ) is approximately 1 in the interface where the cortex resides , ηm is the viscosity coefficient around the cortex region and ηc is the viscosity coefficient for the cytoplasm . We assume that the viscosity around the cortex is proportional to actomyosin concentration , and the viscosity coefficient in the cortex is greater than the one in the cytoplasm ( ηm > ηc ) . The second term , ξ u , is the hydrodynamic drag [25] . We consider the hydrodynamic drag force since the extra-embryonic matrix ( EEM ) , the space between membrane and eggshell , is fluid-filled [5 , 26] . The last term in Eq ( 7 ) , Fmem , is the sum of forces by the membrane/cortex . Since we assume cortex and membrane to be a single entity , the forces resulting from membrane deformations and cortex contractions are considered together as follows: F m e m = F t e n s i o n + F b e n d i n g + F e g g s h e l l + F v o l u m e + F a c t o m y o . ( 8 ) In particular , the first four forces can be derived from the surface free energy E by taking the variational derivative of the energy functional - δ E ( ϕ ) δ ϕ . In the following we describe the individual forces in detail . The domain of our model is taken to be [−45 μm , 45 μm] × [−45 μm , 45 μm] on the x − y plane . Considering the size of an early C . elegans embryo which is approximately 50 μm in length and 30 μm in diameter , the initial phase field function ϕ is taken to be an ellipse with radii 15 and 25 , and is defined by: ϕ ( x , y ) = 0 . 5 ( tanh ( 1 - ( x 15 ) 2 + ( y 25 ) 2 0 . 025 ϵ ) + 1 ) , ( 9 ) where ϵ is the parameter that scales the width of the interface in the phase field function ϕ , and is taken to be 2 in this work . The boundary conditions for ϕ are taken to be periodic boundary conditions for the ease of using Fourier transform in computation . Initially , the concentration of anterior Par proteins is high in the anterior part of the cell and low in the posterior part , and the posterior Par proteins have a reciprocal distribution . To determine initial conditions of the species , the bistable steady-state solutions of model ( 1 ) – ( 5 ) with high and low concentrations are found for all species . The initial distributions of [Am] , [Asd] , [Add] and [P] are defined on the x − y plane and are of the form c 1 ( 0 . 5 - 0 . 5 tanh ( - y - K 0 0 . 5 ϵ ) ) + c 2 , where K0 determines the location of the initial interface between the high and low Par protein concentrations , and c1 and c2 are the high ( low ) and low ( high ) base concentrations . We chose the initial interface between high and low anterior Par proteins to be close to the posterior domain to mimic the initial cue for polarity establishment . The values of K0 , c1 and c2 for each species are given in S1 Table . The initial actomyosin [M] is assumed to be uniformly high . In this work , we take 0 . 7 for the initial actomyosin concentration . No-flux boundary conditions for these species are taken on the membrane . The non-dimensionalized version of model ( 1 ) – ( 5 ) is used in the simulation , and the equations are shown in S1 Text , Eqs . ( S1 ) - ( S5 ) . The associated parameter values listed in S1 Table . are chosen to produce desired model behavior . For the membrane evolution Eqs ( 6 ) – ( 8 ) , the values of all of the parameters , their physical meanings and references are listed in S2 Table . While experimental measurements for the tension and forces are unavailable for this organism , we chose the parameters so that the results qualitatively reproduce the behavior of a pseudocleavage furrow in a wild type C . elegans embryo , which refers to a contracting anterior domain and an invaginating pseudocleavage furrow moving toward the anterior pole as the cell polarizes . To test how robust the model behavior is with respect to parameters , we have perturbed the some parameters in the phase field model by 20% and found that the results are very similar ( see S1 Fig ) . In the numerical simulations , Eqs . ( S1 ) - ( S5 ) are coupled to Eqs ( 6 ) – ( 8 ) alternately: given a fixed membrane profile , the reaction-diffusion equations are solved , followed by solving the phase field model using a fixed protein profile . The details of the implementation of the numerical algorithms are provided in S2 Text .
We first wished to determine if our model could reproduce wild type behavior of the embryo during the establishment phase , as described in the Introduction and Model setup and parameters . In particular , we wish to see if our model can produce a pseudocleavage furrow , an invagination that advects with the edge of the actomyosin cap and the interface between the anterior and posterior Par proteins . This process is illustrated in Fig 1A , and a DIC image of a typical wild type C . elegans embryo at the end of the establishment phase , when the pseudocleavage furrow has reached the middle of the cell , is shown in Fig 1B . Numerical simulations of our model successfully reproduce the wild type behavior , by using the estimated parameters in S1 and S2 Tables . In Fig 2A , we show the distributions of actomyosin ( m , non-dimensionalized [M] ) along with the deforming membrane , and its relative position with respect to the eggshell . The overall profile of the total anterior Par proteins ( a1+ a10+ a11 , non-dimensionalized [Am]+ [Asd]+ [Add] ) is almost identical to that of [M] qualitatively and therefore is not shown here . From Fig 2A , we observe the initial formation of a shallow invagination at the interface between anterior and posterior Par proteins occurs before T = 1600s ( the time of the appearance of the furrow is related to the preset initial cue ) , and as it moves toward the anterior end , the pseudocleavage furrow deepens and reaches the middle of the cell around T = 4600 . Although we took a specific set of parameter in the simulation to demonstrate the wild-type behavior , our numerical experiments showed that it is possible to generate qualitatively similar behavior for a wider range of values for these parameters , which we discuss in The relationship between pseudocleavage furrow depth , time of polarity establishment and actomyosin related forces . It is known that the eggshell provides structural support for the embryo , but details about the mechanical interactions between the eggshell and the embryo are still unclear . In the experiments of [28 , 29] , the authors showed that mutations causing a loss of peri-vitelline space or extra-embryonic matrix ( EEM ) , which is the space between eggshell and membrane , results in embryos lacking a pseudocleavage furrow and unable to initiate cytokinesis . This led us to investigate with our computational model how cell polarization and formation of the pseudocleavage furrow changes under different eggshell-to-membrane distances . In the following , we denote the distance between the eggshell and the membrane by ed . The value of ed is constant throughout each simulation and does not vary as the cell undergoes shape changes . All parameters except ed remained the same as in the wild-type simulations of Fig 2A . In our model , the key element that generates the cell shape changes is the actomyosin contractility forces Factomyo , which consists of two parts: cross-linked actomyosin contractility proportional to the actomyosin concentration ( controlled by the parameter cm ) , and the aligned actomyosin contractility which is proportional to the gradient of the actomyosin concentration ( controlled by cg ) . As described in Mathematical model , we differentiate these two forces by how the actomyosin are structured , although both forces are due to actomyosin contractility . We would like to investigate how modulating cg and cm gives rise to different membrane behaviors during the establishment phase . In particular , one of the aims is to find the ideal ranges for those parameters to generate wild-type behavior qualitatively and understand how these parameters affect the depth of the pseudocleavage furrow and the time of polarity establishment . In this section , we uniformly sampled 11 of cg values and 24 of cm values from the ranges cg ∈ [300 , 800] , cm ∈ [10 , 240] and performed simulations for each set of ( cg , cm ) . We also wished to investigate how the eggshell affects Par protein polarization and cell morphology . This is difficult to study experimentally since the eggshell is essential for the survival of the early embryo . Fortunately , modeling has no such constraint . Here we used the parameter set that produced the wild-type behavior in Fig 2 but take Ms , the coefficient of the eggshell force , to be zero to model the removal of eggshell . As shown in Fig 5A , we observed that without an eggshell , the asymmetric contraction of actomyosin , which is higher in the anterior domain , led to an expanded posterior domain and a small anterior domain as the pseudocleavage furrow moves to the anterior . Moreover , the pseudocleavage furrow reaches the middle of the cell in much shorter time ( 2100 s ) compared to the wild-type case ( 4600 s ) . This suggests that the eggshell might be significant for proper morphology as well as timing of cell polarization . In the above numerical experiment , it was assumed that the presence of the eggshell , a nonzero value of Ms , is the only parameter that is changed in the with/without eggshell cases . However , it is possible that the removal of the eggshell also leads to changes in the aligned actomyosin contractility ( controlled by the parameter cg ) . It may be that compression between the membrane and eggshell exerts extra force , deepening the furrow ingression caused by the aligned actomyosin contractility . If this is the case , then the aligned actomyosin contractility ( cg ) will be smaller when the eggshell is removed . To test this scenario , we decrease cg from 500 ( wild type ) to 300 and take Ms = 0 ( no eggshell ) , and the results are displayed in Fig 5B . Due to the reduced cg , the invagination at the anterior-posterior Par protein interface disappears , that is , there is no pseudocleavage furrow . Similar to Fig 5A , the cell expands in the posterior domain , and despite the loss of furrow , the change of curvature in the cell membrane is still visible because of the differential actomyosin contraction in the anterior and posterior domains . A systematic study of varying cg values in the absence of eggshell showed the continuous change in cell morphology ( S2 Fig ) , and in all cases the time for the anterior-posterior Par protein interface to travel to the middle of the cell is significantly shorter than that for a wild-type cell . Altogether , our simulations in Fig 5 suggest that if the eggshell is removed , the cell shape will be highly distorted , and in some cases , the pseudocleavage furrow may not form . These results suggest that the eggshell might be effective in reinforcing polarization and preserving cell morphology , and the presence of the rigid eggshell may also significantly slow down the polarization .
In this paper , we have used a mathematical model to investigate morphological changes of the C . elegans embryo during the establishment phase of its polarization process . In particular , we are interested in the role of the eggshell in formation of the pseudocleavage furrow , and the interaction between the eggshell and the asymmetric distribution of actomyosin concentration observed during polarization . Our model not only allowed us to qualitatively reproduce the experimentally observed wild-type membrane behavior and pseudocleavage furrow , but also provided biological insights into some scenarios that are unattainable with current experimental tools . Our model combines the well established phase field model to describe the morphogenesis of the cell membrane/cortex by incorporating force generated by several mechanisms: actomyosin contractility in cross-linked and aligned form , constraint from volume conservation , and constraint from the eggshell . The phase field model is coupled with protein dynamics on the cell membrane . Using our mathematical model , we have demonstrated that cell mechanics and geometry may affect protein dynamics on the cell membrane . Previous experiments have shown that cells with mutations that eliminate the space between embryo and eggshell do not have pseudocleavage furrows and are defective in cytokinesis [28] . Our numerical simulations produced consistent results: when the space between the eggshell and membrane is eliminated , no pseudocleavage furrow was observed . Therefore the absence of a pseudocleavage furrow in mutants might have a mechanical explanation in that there is no place for a furrow ingression . The unsuccessful cytokinesis in those mutants can be explained by the lack of room for invagination of the membrane , similar to the reason for the absence of a pseudocleavage furrow . Our results also indicate that the size of the space between the eggshell and the membrane might affect the speed of protein polarization: if there is no space between eggshell and embryo , the speed of protein polarization is greatly attenuated , possibly due to the lack of a pseudocleavage furrow . If it would be experimentally possible to increase the space between the embryo and the eggshell , the model predicts that the cell will experience a shorter time to polarization and a deeper furrow relative to a wild type embryo . Our results also indicate a relationship between force generated by cross-linked actomyosin contractility and force generated by aligned actomyosin contractility , a potential area for further experimental investigation . In the numerical experiment in which the eggshell is removed , we observed that a highly asymmetric contraction of actomyosin leads to a distorted cell shape . This result suggests that the mechanical support of an eggshell plays an essential role in proper protein polarization and that if the eggshell can be removed while maintaining cell integrity , the model predicts large scale morphological changes and a small anterior domain relative to the posterior . In Mayer et al . [30] , the authors found that , in the anterior , the tension in the direction orthogonal to the anterior-posterior axis is different than the tension along the anterior-posterior axis . The anisotropy of tension increases the depth of the pseudocleavage furrow [27] . In this work , our aim was to investigate the qualitative behavior of the pseudocleavage furrow by taking advantage of a 2D model in the phase field context , therefore the simplest form of tension was considered . Another possible mechanism not being considered in this work is tension generated by compression in the tangential direction . It is possible that the eggshell increases compression , and beyond some point the cortex-membrane complex might generate the pseudocleavage furrow as a result of a buckling type of instability , leading to formation of a crease [31] . In this case , removal of the eggshell might reduce the compression , which will likely lead to the absence of a pseudocleavage furrow . However , to understand buckling behavior due to compression , we would need to include advection of the cortex in the tangential direction which cannot be tracked by the boundary tracking method we use . In this investigation , we focussed on the large invagination formed by the pseudocleavage furrow . However , the early embryo also exhibits shallow , transient invaginations called ruffles during the establishment phase . The mechanisms behind ruffle formation are not clear . Limited change in area , contractility by actomyosin and the eggshell may have a combined effect leading to formation of creases on the surface , creating ruffles . In our model , we assumed that area can vary due to high tension in the anterior . Our model captures the overall area change in the anterior due to contraction without including the self-intersecting and invaginating area seen with ruffles . A more detailed model including cross-linking actomyosin foci might be needed to explain the mechanism behind the generation of both the pseudocleavage furrow and ruffles which we leave for future work .
|
Polarization , whereby molecules and proteins are asymmetrically distributed throughout the cell , is a vital process for many cellular functions . In the early C . elegans embryo the asymmetric distribution of cell cytoskeleton during the initiation of polarization leads to asymmetric contractions which are higher in the anterior and lower in the posterior of a cell . The C . elegans embryo is surrounded by a rigid body , the eggshell , which functions in numerous cell processes . We investigate the structural support of eggshell during the establishment phase by tracking the moving cell surface . We incorporate protein dynamics involved in polarization into the membrane evolution . We conclude that eggshell might have a role in cell polarization by preventing the distortion of cell surface .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"classical",
"mechanics",
"fluid",
"mechanics",
"caenorhabditis",
"surface",
"tension",
"animals",
"animal",
"models",
"membrane",
"proteins",
"developmental",
"biology",
"caenorhabditis",
"elegans",
"viscosity",
"model",
"organisms",
"materials",
"science",
"experimental",
"organism",
"systems",
"damage",
"mechanics",
"materials",
"physics",
"embryos",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"bending",
"chemical",
"properties",
"physical",
"chemistry",
"embryology",
"deformation",
"chemistry",
"cell",
"membranes",
"continuum",
"mechanics",
"physics",
"biochemistry",
"biochemical",
"simulations",
"eukaryota",
"cell",
"biology",
"nematoda",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"organisms"
] |
2018
|
The importance of mechanical constraints for proper polarization and psuedo-cleavage furrow generation in the early Caenorhabditis elegans embryo
|
Natural killer ( NK ) cells classically typify the nonspecific effector arm of the innate immune system , but have recently been shown to possess memory-like properties against multiple viral infections , most notably CMV . Expression of the activating receptor NKG2C is elevated on human NK cells in response to infection with CMV as well as HIV , and may delineate cells with memory and memory-like functions . A better understanding of how NKG2C+ NK cells specifically respond to these pathogens could be significantly advanced using nonhuman primate ( NHP ) models but , to date , it has not been possible to distinguish NKG2C from its inhibitory counterpart , NKG2A , in NHP because of unfaithful antibody cross-reactivity . Using novel RNA-based flow cytometry , we identify for the first time true memory NKG2C+ NK cells in NHP by gene expression ( KLRC2 ) , and show that these cells have elevated frequencies and diversify their functional repertoire specifically in response to rhCMV and SIV infections .
Although NK cells have traditionally been thought to be innate immune cells that lack the antigen-specificity seen in the adaptive immune system , NK cells have very recently been reported to possess memory and memory-like functions [1–8] . Though this area of investigation is currently developing , subpopulations of NK cells that express NKG2C ( CD159C ) in humans or Ly49H and Ly49P in mice mobilize in response to CMV infection [9–13] . While this phenomenon has been described in human and murine studies , because of technical limitations it has not yet been possible to examine memory and memory-like NKG2C+ NK cells in NHP models . This is predominantly attributed to the high degree of homology in NHP between the extracellular domains of two NKG2 isoforms , activating NKG2C and inhibitory NKG2A –making the two indistinguishable via currently available antibodies and standard measurements [14 , 15] . NHP models are crucial to multiple areas of medical research , including HIV and CMV infectious disease study and transplant biology [16–18] since the murine system does not always approximate human immunology . As such , the inability to study NKG2C+ memory NK cells in NHP models remains a major research deficit . NKG2C and NKG2A both belong to the C-type lectin family of NK cell receptors . NKG2C recruits the adaptor protein DAP12 , which has an ITAM ( immunoreceptor tyrosine-based activation motif ) , and NKG2A has two ITIM ( immunoreceptor tyrosine-based inhibitory motif ) domains , which lead to recruitment of phosphatases , and downregulation of signaling [19 , 20] . Because these two proteins act in opposition to each other , it is crucial to discriminate between cells that express either protein in order to more accurately determine what role these cells play during infection . As a result , we aimed to utilize RNA hybridization technology recently adapted for flow cytometry ( PrimeFlow ) to label the gene transcripts of rhesus macaque NKG2A and NKG2C ( KLRC1 and KLRC2 , respectively ) , taking advantage of several nucleotide differences between the two transcripts , in order to distinguish cells that transcribed these isoforms . This approach should allow simultaneous detection of surface and intracellular proteins as well as gene transcript levels with a single-cell resolution using polychromatic flow cytometry . In addition to differentiating between KLRC1+ and KLRC2+ NK cells , this technology should allow evaluation of NK cell population diversity , including memory cells , in the context of “normal” CMV infection , in chronic SIV infection , and in CMV-negative specific pathogen free ( SPF ) rhesus macaques . Understanding how KLRC1±KLRC2± NK cells function in the context of infection will help improve our basic understanding of NK cell biology , potentially inform preclinical HIV vaccine or cure studies relying on macaque models , and provide a significant technological advance to the study of memory NK cells in primates .
All animals were housed at the Tulane Primate Research Center ( TNPRC ) or Biomere ( Worcester , MA ) . All study blood samplings were reviewed and approved by the Tulane University Institutional Animal Care and Use Committee or the Biomere Institutional Animal Care and Use Committee under protocol numbers 16–08 and 17–02 . All animal housing and studies were carried out in accordance with recommendations detailed in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health with recommendations of the Weatherall report; “The use of non-human primates in research” . AAALAC numbers for TNPRC and Biomere– 00594 and 1152 , respectively . Animals were fed standard monkey chow diet supplemented daily with fruit and vegetables and water ad libitum . Social enrichment was delivered and overseen by veterinary staff and overall animal health was monitored daily . Animals showing significant signs of weight loss , disease or distress were evaluated clinically and then provided dietary supplementation , analgesics and/or therapeutics as necessary . No animals were euthanized as part of this research . Thirty Indian rhesus macaques were analyzed in this study: ten specific pathogen free ( SPF ) animals ( rhCMV- and SIV-negative ) , twelve otherwise naïve animals that were naturally infected with rhCMV ( rhCMV+ ) , and eight chronically infected with SIVmac251 ( all of which rhCMV+ ) . SPF animals and age-matched non-SPF/rhCMV+ macaques were housed at the Tulane National Primate Research Center ( TNPRC ) . SIV-infected macaques and additional rhCMV+ animals were housed at Biomere . All animals were colony housed until on study and then infected animals were housed under BSL2 conditions . Whole blood was collected into EDTA-treated tubes . Peripheral blood mononuclear cells ( PBMCs ) were isolated by density-gradient centrifugation layered over 100% Ficoll . Cell aliquots were immediately analyzed or cryopreserved in 90% FBS , 10% DMSO ( Sigma ) and stored in liquid nitrogen vapor . PBMCs were thawed and rested for 12h in R10 media at 37°C prior to surface and intracellular staining followed by RNA-Flow hybridization using the manufacturer’s recommended protocol ( PrimeFlow , Affymetrix , Santa Clara , CA ) with the antibodies detailed in the flow cytometry section below , and with rhesus macaque-specific KLRC1 and KLRC2 probesets . Rhesus-specific probesets were custom designed with the assistance of Affymetrix ( Santa Clara , CA ) specifically for this project to target rhesus KLRC1 and KLRC2 . Target probes sequences for KLRC1 and KLRC2 are shown in S1 and S2 Figs , as are ‘blocking probe’ sequences used to prevent nonspecific binding . The blocking probes were designed in order to avoid amplifying/detecting undesired NKG2 homologues . Blocking probes do not have the ability to form branched DNA structures which hybridize to the label probe fluorophores as opposed to the target probes which are able to hybridize to label probe fluorophores . Both target probes and blocking probes were simultaneously added at the target probe hybridization step as per the manufacturer’s protocol . Probesets were labeled with Alexa-488 ( KLRC1 ) and Alexa-647 ( KLRC2 ) fluorophores by Affymetrix ( Santa Clara , CA ) . All KLRC1 and KLRC2 gates were determined for each sample by comparing the samples stained with all antibodies and probesets with samples only stained with antibodies ( no probeset control ) . All antibodies used were purchased from BD Biosciences unless specified otherwise . For phenotypic panels antibodies against the following cell antigens were used: CD2 ( RPA2 . 10 ) , CD3 ( SP34 . 2 ) , CD337 ( p30-15 ) , CD14 ( MϕP9 ) , CD20 ( L27 ) , CD16 ( 3G8 ) , CD56 ( NCAM16 . 2 ) , HLA-DR ( G46-6 ) , CD8α ( SK1 ) , KIR2D ( NKVFS , Miltenyi [this antibody recognizes KIR3D in NHP as shown by Pomplun , N . et al . [21]] ) , CD159a/c ( Z199 , Beckman Coulter ) , CD366 ( F38-2E2 , Biolegend ) . Additionally , antibodies used for functional assays included TNF-α ( MAb11 ) , IFN-γ ( B27 ) , CD107a ( H4A3 ) . Flow cytometry data was acquired on a LSRII ( BD Biosciences , La Jolla , CA ) and analyzed with FlowJo software ( version 10 . 2 , Tree Star , Ashland , OR ) . t-SNE ( t-distributed stochastic neighbor embedding ) was carried out using the t-SNE feature in FlowJo using 1000 iterations and a perplexity of 20 . PBMCs from animals were prepared at 37°C and cells were cultured with Golgi Plug and Golgi Stop ( BD Biosciences , concentrations as recommended by manufacturer ) , and PMA ( 3 . 3μg/mL , Sigma ) and Ionomycin ( 5μg/mL , Sigma ) or with unlabeled anti-CD16 ( 3G8 , 20μg/mL ) and cross-linked with F ( ab’ ) 2 ( 20μg/mL , Jackson Immunoreserach , West Grove , PA ) for 14h in R10 media ( RPMI + 10% FBS + 2% PenStrep ( Gibco ) ) . Rhesus plasma was assessed for rhCMV-specific IgG by a previously reported rhCMV UCD52 whole virion ELISA [22] . After plates were coated with 4 , 400 PFU/mL of rhCMV UCD52 virus , the previously reported procedure was followed . The positivity threshold for detectable antibody levels was set to equal twice the OD of a rhCMV-seronegative plasma control at the starting dilution ( 1:30 ) . Statistical and graphing analyses were performed with GraphPad Prism 7 . 0 software ( GraphPad Software , La Jolla , CA ) . Nonparametric Mann-Whitney U tests and Wilcoxon tests were used where indicated , and a p-value of p < 0 . 05 was considered to be statistically significant .
Total NK cells were identified among PBMC in rhesus macaques using traditional phenotypes optimized by our laboratory [5 , 23–27]: CD3-CD14-CD20-NKG2A+ ( Fig 1A ) . Unsurprisingly , the anti-NKG2A antibody was unable to distinguish between NKG2A and NKG2C , as has been previously shown by the Letvin lab whereby they showed that anti-human NKG2A antibodies were cross-reactive with four NKG2C alleles [14] . As a result we now classify bulk cells that are positive for this antibody as NKG2AC+ NK cells . Using RNA-Flow ( see Methods ) we next identified within NKG2AC+ NK cell populations those cells that expressed transcripts of the genes coding for NKG2A and NKG2C ( KLRC1 and KLRC2 , respectively ) , and accurately quantified the frequency of NK cells expressing one or both of these genes ( Fig 1A–1C ) . Interestingly , absolute frequencies of both KLRC1+ and KLRC2+ NK cells ( Fig 1B ) were lower in SPF animals compared to either rhCMV+ or SIV-infected macaques— ( KLRC1 ) 0 . 18% , 0 . 29% , and 0 . 68% of lymphocytes in SPF , rhCMV+ , and SIV+ animals respectively; and ( KLRC2 ) 0 . 31% , 2 . 55% , and 2 . 04% of lymphocytes in SPF , rhCMV+ , and SIV+ animals respectively . These data demonstrate that while NK cells are less frequent in SPF animals in general , following rhCMV infection a less than 2-fold non-significant increase occurs in KLRC1+ NK cells , but the increase in KLRC2+ NK cells is 12-fold . This dramatic observation is congruent with other findings in human research , which show elevation of NKG2C+ NK cells specifically following CMV infection [9–11 , 28] . Examining the frequency of KLRC1+ and KLRC2+ NK cells relative to the total NK cells population ( Fig 1C ) also revealed that in rhCMV+ and SIV-infected macaques there was an obvious reduction in KLRC1+ NK cells in lieu of KLRC2+ NK cells , but surprisingly , expression of both KLRC1 and KLRC2 remained high in NK cells from SPF animals . To further clarify our findings , we re-optimized our technical approach to measure both KLRC1 and KLRC2 simultaneously . Using the gating strategy shown in Fig 2A we were able to clearly distinguish four distinct NK cell populations by expression of KLRC1 and KLRC2 . Strikingly , this analysis revealed that in the absence of rhCMV infection , a KLRC1+KLRC2+ population is dominant ( Fig 2B ) . In contrast , in rhCMV+ animals ( rhCMV+ and SIV+ groups ) the predominant population was single-positive KLRC1-KLRC2+ . Consistently , KLRC1+KLRC2- and KLRC1-KLRC2- represented minority populations among all animal groups and could represent precursor or aberrant NK cells outside the normal NK cell repertoire . While the presence of the KLRC1-KLRC2- population was surprising , it must be noted that the Z199 clone that detects human NKG2A and rhesus macaque NKG2A and NKG2C is promiscuous and could be identifying minor NKG2 isoforms[14] . The extreme specificity and blocking probes used in the RNA-based flow cytometric technique make it highly unlikely to have non-specific signals ( S1 and S2 Figs ) . The presence of a double-negative population is more likely resulting from some samples where mRNA levels being below the threshold of detection of this assay . Regardless , these findings point to the overall importance of this study which are now able to confirm true KLRC2+ ( NKG2C ) NK cells in macaques which have only been incompletely described previously . Also , consistent with observations in humans we also find that both KLRC1 and KLRC2 are expressed on minor populations of T cells ( S3 Fig ) . Finally , we can determine that the observation that rhCMV+ and SIV+ animals have higher relative and absolute frequencies of KLRC1-KLRC2+ compared to KLRC1+KLRC2+ NK cells is likely analogous to the memory and memory-like functions observed in human CMV infection [10 , 12 , 13] , whereby prior to CMV exposure both inhibitory NKG2A and activating NKG2C are expressed , but NKG2A is downregulated following infection . Further supporting the notion that CMV specifically expands NKG2C+ NK cells , we found a significant positive correlation between increasing KLRC1-KLRC2 ( NKG2C ) + NK cells and rhCMV-binding IgG , as a surrogate indicator of virus replication ( Fig 3D ) . Concurrently there was a significant negative correlation between frequencies of KLRC1+KLRC2± NK cells and increasing rhCMV-specific IgG ( Fig 3A and 3B ) . There was , however , no association between rhCMV-specific IgG and KLRC1-KLRC2- NK cells ( Fig 3C ) . Interestingly , no correlation was found between SIV viral loads ( median 3 . 00 x 106 virus copies/ml; range 5 . 66 x 104 to 3 . 30 x 107 ) and any of the NK cell subpopulations . Collectively , these data suggest that perturbations of KLRC1±KLRC2± NK cells is primarily driven by rhCMV status . We next wanted to confirm phenotypically that the KLRC1 and KLRC2 definitions were indeed identifying NK cell subpopulations that are analogous to their human counterparts in which NKG2C+ NK cells are more activated and differentiated . Indeed FcRγIII receptor CD16 was higher in rhCMV+ and SIV-infected animals compared to SPF and was consistently higher on KLRC2+ NK cells . ( Fig 2C , S1 Table ) . CD56 is typically expressed on most circulating NK cells in humans , but is expressed on only a small frequency of cytokine-producing or less differentiated NK cells in macaques [29] . Consistent with this notion , we found CD56 expression was higher in general on homeostatic KLRC1+KLRC2+ NK cells , but was poorly expressed on KLRC1-KLRC2+ NK cells expanded by rhCMV infection seen in the CMV+ and SIV+ groups . While this trend was not present in SPF animals we noted that expression of CD56 was generally higher in SPF samples as compared to the CMV+ and SIV+ groups , though it was not statistically significant . In general KIR expression is increased as NK cells differentiate and it was hypothesized that rhCMV or SIV infection are increasing activation and differentiation . Indeed expression of the common macaque KIR , KIR3D , was lower in all SPF animals and had the highest expression on KLRC1-KLRC2+ NK cells . CD2 , which has been shown to synergize with NKG2C to promote adaptive NK cell functions [30] , was also found to be increased on KLRC2+ cells in rhCMV+ macaques . Unfortunately , cross-reactive antibodies against the CD57 carbohydrate epitope , also associated with memory NK cell phenotypes , do not currently exist for monkeys and thus could not also be evaluated here . Nonetheless , these findings collectively suggest this population is generally more activated and differentiated and has a phenotype consistent with adaptive functions . Since our analyses suggested rhCMV infection may be driving KLRC2+ NK cell expansion , we next used t-SNE to evaluate NK cell subpopulation clustering and diversity . NK cells from SPF animals clustered into two major groups–corresponding with KLRC1+KLRC2+ and KLRC1-KLRC2+ populations . In contrast , NK cells from rhCMV+ and SIV+ animals clustered into far more minor and distinct groups ( Fig 4A ) . The phenotypic characteristics of these groups were also highly variable depending on infection status and subpopulation ( Fig 4B ) . These data suggest that there is a greater diversity of NK cell subpopulations following infection with rhCMV or SIV as compared with the uninfected SPF group . These findings are in strong agreement with previous analyses by showing that human NK cell diversity increases following infection with HIV and other pathogens [31–33] . Next we wanted to examine whether there were any functional differences among each of the KLRC1+ and KLRC2+ populations . Mitogen stimulation revealed that all NK cell subsets from all animal groups were capable of surrogate cytotoxic ( CD107a ) and cytokine-based ( IFN-γ , TNF-α ) responses ( Fig 5 ) . Unfortunately , the number of positive events for the KLRC1+KLRC2- population were too few for us to reliably report from the functional assay . Nevertheless , the remaining quadrant populations were functional , with the KLRC1+KLRC2+ and KLRC1-KLRC2+ populations demonstrating the most robust responses . Interestingly , upon mitogen stimulation we observed that NK cells from SPF animals produced proportionately greater cytokines , which could be indicative of a more immature status as expected given the lack of virus exposure . Following stimulation through CD16 that could mimic ADCC , again all NK cell subpopulations were functionally competent . Perhaps even more obvious in this assay , NK cells from SPF animals favored cytokine production , whereas those from rhCMV+ animals were more adept for CD107a upregulation as a surrogate indicator of cytotoxicity ( S4 Fig ) . These findings further suggest that CMV infection is necessary to activate and prime cytotoxic functions , particularly those dependent on antibodies and corroborates findings in humans mediated by NKG2C+ γ-chain deficient memory-like NK cells [13] . Indeed , the memory-like programming observed for rhCMV could be epigenetic in nature if not memory per se . Collectively , all NK subpopulations from SIV-infected animals were functionally responsive to mitogen , but were poorly responsive to CD16 cross-linking . These findings are well in-line with previous observations of NK cell dysfunction in HIV and SIV infections . In this paper we present a cross-sectional analysis of several infected and uninfected animals . Further studies need to be carried out in order to examine the kinetics of infection and how SIV or CMV may play a role in shaping the NK cell repertoire . Importantly , we also show that with this technique we are able to also examine and identify KLRC1 and KLRC2+ NK cells from peripheral lymphoid tissues such as the spleen and primary sites of infection such as the colon ( S5 Fig ) . This will allow us to examine the role that these NK cell populations play in the earliest stages following infection in the relevant tissues . In conclusion , we report that it is now possible to specifically identify NKG2C+ and NKG2A+ macaque NK cells using their respective transcripts , KLRC2 and KLRC1 , as proxy . Further , we show for the first time that rhCMV infection results in increased NK cell diversity and a specific increase in NKG2C+ NK cells . Altogether these findings strengthen the argument for NKG2C+ memory and memory-like NK cells arising in response to CMV and lentivirus infections and provide a tangible NHP model in which to study them .
|
Natural killer ( NK ) cells are a crucial component of the early innate immune response , and although NK cell responses have been thought be only non-specific , recent evidence suggests that NK cells are capable of expanding with some specificity , indicative of a memory-like adaptive response . The activating receptor NKG2C has been one cell surface protein associated with this memory-like NK cell expansion in the context of CMV and HIV infection in humans , yet very little is known about NKG2C+ NK cells in non-human primate ( NHP ) animal models . This is predominantly because there are no antibodies that can distinguish NKG2C from other NKG2 family molecules in NHP . Because vaccine and cure-related studies for HIV rely heavily on NHP models , this is a significant impediment towards understanding an NK cell population that may possibly improve responses to HIV . In this paper we present a solution , by adapting a technique whereby mRNA specific to NKG2C and NKG2A ( KLRC2 and KLRC1 , respectively ) is fluorescently labeled while the cell is simultaneously stained using traditional flow cytometry , and provide a first-ever characterization of NKG2C+ NK cells in NHP . Further , we show that NKG2C+ NK cells expand in a memory-like fashion following rhCMV and SIV infections .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"vertebrates",
"animals",
"mammals",
"cell",
"differentiation",
"retroviruses",
"primates",
"immunodeficiency",
"viruses",
"viruses",
"animal",
"models",
"developmental",
"biology",
"rna",
"viruses",
"experimental",
"organism",
"systems",
"immunologic",
"techniques",
"old",
"world",
"monkeys",
"research",
"and",
"analysis",
"methods",
"rhesus",
"monkeys",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"monkeys",
"animal",
"cells",
"medical",
"microbiology",
"microbial",
"pathogens",
"immunoassays",
"spectrophotometry",
"macaque",
"siv",
"cytophotometry",
"eukaryota",
"cell",
"biology",
"nk",
"cells",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"lentivirus",
"amniotes",
"organisms"
] |
2018
|
Tracking KLRC2 (NKG2C)+ memory-like NK cells in SIV+ and rhCMV+ rhesus macaques
|
The murine leukaemia virus ( MLV ) Gag cleavage product , p12 , is essential for both early and late steps in viral replication . The N-terminal domain of p12 binds directly to capsid ( CA ) and stabilises the mature viral core , whereas defects in the C-terminal domain ( CTD ) of p12 can be rescued by addition of heterologous chromatin binding sequences ( CBSs ) . We and others hypothesised that p12 tethers the pre-integration complex ( PIC ) to host chromatin ready for integration . Using confocal microscopy , we have observed for the first time that CA localises to mitotic chromatin in infected cells in a p12-dependent manner . GST-tagged p12 alone , however , did not localise to chromatin and mass-spectrometry analysis of its interactions identified only proteins known to bind the p12 region of Gag . Surprisingly , the ability to interact with chromatin was conferred by a single amino acid change , M63I , in the p12 CTD . Interestingly , GST-p12_M63I showed increased phosphorylation in mitosis relative to interphase , which correlated with an increased interaction with mitotic chromatin . Mass-spectrometry analysis of GST-p12_M63I revealed nucleosomal histones as primary interactants . Direct binding of MLV p12_M63I peptides to histones was confirmed by biolayer-interferometry ( BLI ) assays using highly-avid recombinant poly-nucleosomal arrays . Excitingly , using this method , we also observed binding between MLV p12_WT and nucleosomes . Nucleosome binding was additionally detected with p12 orthologs from feline and gibbon ape leukemia viruses using both pull-down and BLI assays , indicating that this a common feature of gammaretroviral p12 proteins . Importantly , p12 peptides were able to block the binding of the prototypic foamy virus CBS to nucleosomes and vice versa , implying that their docking sites overlap and suggesting a conserved mode of chromatin tethering for different retroviral genera . We propose that p12 is acting in a similar capacity to CPSF6 in HIV-1 infection by facilitating initial chromatin targeting of CA-containing PICs prior to integration .
The retroviral Gag polyprotein plays essential roles at multiple stages of the viral life cycle . In late infection , Gag mediates the assembly and release of progeny viral particles . It is then proteolyically cleaved during viral maturation into a number of individual proteins . In addition to the three main structural proteins , matrix ( MA ) , capsid ( CA ) and nucleocapsid ( NC ) , many retroviruses also produce additional Gag cleavage products such as the p12 protein of MLV and the p6 protein of HIV-1 [1] . The proline-rich late ( L ) -domains which allow Gag to recruit the cellular ESCRT machinery required for efficient virus budding are frequently found in these additional Gag cleavage products ( Fig 1A ) [2–5] . However , it is likely that these proteins also have other functions , for example , MLV p12 is known to be essential for early replication events . Initially , a subset of alanine-scanning substitution mutations in Moloney ( Mo ) -MLV p12 were shown to inhibit replication of MLV , predominantly at a stage post reverse transcription [6] . These mutations clustered into two regions on p12 , one located towards the 5’ and the other towards the 3’ end of the gene . These two groups of mutants were later shown to be phenotypically distinct from each other and two functional domains within the protein were defined , both of which were conserved among p12 orthologues of other gammaretroviruses [7] . Furthermore , infections with mixed particle containing both wild type ( WT ) and mutant viruses revealed that the two domains act sequentially in the virus life cycle , with the N-terminal function required before that of the C-terminus [7] . Functional cooperativity between p12 and CA was first demonstrated in a study of chimeric MLV/ spleen necrosis virus ( SNV ) particles , with the chimeras being infectious only when p12 ( p18 in the case of SNV ) was accompanied by CA from the same virus [8] . Recent work from our laboratory has shown that the N-terminal domain ( NTD ) of p12 ( aa 10–30 of Mo-MLV p12 ) binds to and stabilises the CA lattice [9] . Alterations in p12 do not affect Gag processing , but viral particles carrying mutations in the NTD of p12 ( mutants 5–8 in Fig 1A ) display core abnormalities and reduced binding to CA-targeting host restriction factors [7 , 9] . The precise mechanism by which p12 influences global core stability has not yet been established . In recent years , a potential role for the p12 C-terminal domain ( CTD ) ( aa 60–74 of Mo-MLV p12 ) in tethering the pre-integration complex ( PIC ) to mitotic chromatin has emerged . Mo-MLV p12 has been observed to co-localise with viral DNA in infected cells using fluorescence microscopy [10] . Viral complexes containing p12 initially localise in the cytoplasm , then traffic towards the nucleus and accumulate on mitotic chromosomes [10 , 11] . Live cell imaging of GFP-p12 carrying viral particles has further revealed that the association with chromatin is a transient event , with p12 being released upon chromatin decondensation [11] . Importantly , mutant viral particles carrying alterations in the p12 CTD ( mutants 13–15 in Fig 1A ) are deficient in chromatin docking and replication , but are partially rescued in both tethering and infectivity by the insertion of heterologous chromatin binding modules into p12 [7 , 11 , 12] . Tethering to chromatin in mitosis is a nuclear retention mechanism shared between different types of viruses . DNA viruses such as gamma-herpesviruses and papillomaviruses use this strategy to maintain their episomal genomes in the host nuclei during cell division [13] . On the other hand , retroviruses must interact with chromatin in order to integrate their genomes as an essential part of their replication . Some retroviruses , including the gammaretrovirus Mo-MLV and the spumavirus prototypic foamy virus ( PFV ) , can only gain access to chromatin during mitosis when the nuclear envelope breaks down , and may therefore initially tether to mitotic chromatin in order to reside in the nucleus once the nuclear membrane has reformed . Viral proteins that mediate chromatin tethering interact with diverse chromatin components . The EBNA-1 protein of the Epstein-Barr virus ( EBV ) binds both host DNA and EBP2 , a chromatin-associated protein , whereas the LANA protein of Kaposi’s sarcoma herpes virus ( KSHV ) and the chromatin binding sequence ( CBS ) of PFV Gag both bind directly to nucleosomal histones H2A-H2B [14–18] . It is not known whether p12 directly binds chromatin or if it interacts with another chromatin binding factor . Notably , fluorescence microscopy revealed that recombinant Mo-MLV p12 did not localise to mitotic chromatin in mammalian cells [19] . One possibility is that another viral protein is required for chromatin tethering . For example , binding to CA could alter the conformation of p12 and influence its affinity for chromatin , or the viral integrase ( IN ) protein could contribute to chromatin tethering of the PIC via its interactions with the bromodomain and extraterminal domain ( BET ) family of proteins [20–22] . Alternatively , the inability of recombinant Mo-MLV p12 to bind chromatin could be due to the absence of an essential post-translational modification . Previous studies have identified p12 as the main phosphorylated protein in Mo-MLV with the majority of phosphorylation occurring on serine-61 within the CTD [19 , 23–25] . In contrast with viral p12 , recombinant Mo-MLV p12 was found to be non-phosphorylated in mammalian cells [19] . However , the importance of p12 phosphorylation for viral replication is unclear . Substitution of S61 with both phosphoablative alanine and ‘phosphomimetic’ aspartic acid causes severe defects in Mo-MLV infectivity [19 , 25] . Furthermore , viral revertants that arise from live passaging of a p12 mutant , SS ( 61 , 65 ) AA , show near WT levels of infectivity without recovery of phosphorylation [25] . Instead , revertants that maintain the original SS ( 61 , 65 ) AA mutations carry additional changes such as an isoleucine substitution at M63 ( M63I ) or R/K substitutions at various residues [25] . In general , these compensatory mutations localise either within the p12 CTD region or adjacent to it , suggesting a possible involvement in chromatin binding . In fact , the M63I substitution has been observed to rescue mitotic chromatin tethering of SS ( 61 , 65 ) AA in infected cells [11] . Interestingly , M63I can also rescue chromatin binding of recombinant Mo-MLV in mammalian cells , suggesting that it could be compensating for an absence of phosphorylation in this context [19] . In this study , we characterised the interactions of p12 with both viral and host proteins , using a combination of virological , biochemical and imaging assays . We observed a p12 CTD-dependent association of CA with mitotic chromatin in MLV-infected cells , suggesting that CA is still present when p12 tethers the PIC to chromatin . We also showed that although recombinant GST-tagged WT Mo-MLV p12 shows little detectable chromatin association in a range of cell-based assays , other gammaretroviral p12 proteins do show chromatin interactions , and that subtle mutations can increase the association of Mo-MLV p12 with chromatin . Furthermore , we established that recombinant p12 has a higher affinity for chromatin in mitosis correlating with an increase in its phosphorylation . A global mass spectrometry-based analysis of recombinant p12-chromatin interactions , identified significant overlap with the chromatin interactome of PFV CBS . Importantly , using biolayer interferometry , we were able to detect direct binding of several p12 orthologs , including WT Mo-MLV p12 , to purified recombinant nucleosomes , and show that these p12 proteins could compete with PFV CBS for chromatin . This suggests a shared evolutionary mechanism for histone binding between gammaretroviruses and foamy viruses .
We recently demonstrated direct binding of purified N-tropic MLV ( N-MLV ) p12_WT to recombinant CA by utilising an assay that was established for studying the interactions of CA with the restriction factor Fv1 [9 , 26] . In this assay , His-tagged N-MLV CA was immobilised on lipid tubes , comprising Ni-chelating DGS-NTA , to enable it to adopt a regular hexameric arrangement . Although an NTD mutant of p12 did not interact with CA , we did not evaluate p12 CTD mutants for CA binding . Therefore , to investigate the contribution of the p12 CTD in the interaction with CA , we compared the binding of purified N-MLV p12_WT , p12_mut6 ( an NTD mutant , Fig 1A ) and p12_mut14 ( a CTD mutant , Fig 1A ) proteins to CA-coated lipid tubes . Purified p12 was incubated with CA-coated tubes and CA complexes were separated from unbound proteins by centrifugation through a sucrose cushion . The pellets were subsequently probed for p12 and CA by western blotting . In contrast with p12_mut6 , p12_mut14 showed similar binding to WT CA as p12_WT ( Fig 1B , lane 2 ) . Furthermore , neither p12_mut14 nor p12_WT bound P1G CA which does not form regular arrays ( Fig 1B , lane 3 ) . These results therefore suggest that alterations in the CTD of p12 do not significantly affect CA binding . Although purified p12 appears to bind recombinant CA arrays in vitro , an association between p12 and CA in viral particles has not yet been demonstrated . In fact , in a previous study , CA co-immunoprecipitated with p12 from lysates of Mo-MLV infected cells but not from lysates of extracellular virions [11] . To demonstrate an association between p12 and CA in virions , we next performed co-immunoprecipitation ( Co-IP ) assays using lysates from purified virion-like particles ( VLPs ) . Prior to performing the immunoprecipitations , the viral lysates were normalised based on their CA content and treated with 1% formaldehyde . Due to its short cross-linking span ( 2–3 Å ) , formaldehyde is commonly used to facilitate the detection of specific protein-protein interactions [27] . Importantly , an antibody targeting p12 was able to immunoprecipitate CA in our assays ( Fig 1C ) . In comparison with p12_WT ( Fig 1C , lane 2 ) , the amount of CA that co-immunoprecipitated was similar for p12_mut14 ( Fig 1C , lane 6 ) but much lower for p12_mut6 ( Fig 1C , lane 4 ) . The CA shell of the mature MLV virion is composed of a network of hexagonal rings [28] . To identify whether p12 recognises monomeric or hexameric CA , we next probed the binding of purified p12 to His-tagged CA immobilised on either lipid nanotubes or on Ni-NTA beads . Unlike on lipid nanotubes , the CA molecules arrayed on beads would be expected to be randomly-oriented . As shown in Fig 1D , even though the Ni-NTA beads carried similar amounts of CA as the lipid tubes , they did not detectably bind p12 ( Fig 1D , lane 4 ) . Our results therefore suggest that , similarly to CA-binding restriction factors [26] , p12 only recognises hexameric CA . In addition , we used the lipid tube-based binding assay to determine the approximate stoichiometry of the p12-CA interaction ( S1 Fig ) . For this purpose , the binding assays were performed with a 16-fold molar excess of p12 to promote the occupation of all potential p12 binding sites on CA . Comparing the immunoblot band intensities to those of a standard curve of known p12 protein concentrations , we calculated the amount of p12 pelleted by the CA-nanotubes as 11 . 17 pmol . Given that we loaded 67 pmol CA onto the nanotubes , this gives a p12:CA ratio of 1:6 , which is suggestive of an interaction surface which is formed upon CA hexamerisation . However , this stoichiometry of one p12 molecule per CA hexamer may be an under-estimation as some CA molecules immobilised on the tubes may not be present in the exact hexameric arrangement conducive to p12 binding . As the CTD of p12 did not appear to interact with CA , we next asked how alterations in the p12 CTD affected CA localisation during virus replication . We therefore immunostained Mo-MLV-infected cells for p12 and CA simultaneously . Cells were synchronised by treatment with aphidicolin and then released from the G1/S block 30 minutes prior to infection . Cells were stained 10 h post-infection . Interestingly , in contrast to a previous study [11] , CA was observed to co-localise with p12 on mitotic chromatin in cells infected with WT VLPs ( Fig 2A ) . Quantitative analysis of chromatin-associated viral complexes in twelve mitotic cells ( ~45 tethered complexes/cell on average ) revealed that 57±6% of p12-puncta were co-stained for CA and that 68±4% of CA-puncta were co-stained for p12 . In mitotic cells infected with VLPs carrying p12_mut14 , neither p12 nor CA tethered to chromatin ( Fig 2B ) . However , p12 and CA still co-localised in the cytosol of these cells ( 39±4% of p12-puncta co-localised with CA and 54±5% of CA-puncta co-localised with p12 in five mitotic cells ) . Our results therefore suggest that binding of CA to p12 is not disrupted upon p12 CTD-mediated chromatin association , identifying how p12 can act as a tether between host chromatin and the viral PIC . In addition to p12 , MLV also carries another protein known to interact with chromatin , IN . In fact , target site selection for MLV integration is known to be mediated by the binding of IN to chromatin-associated BET proteins [20–22] . How the IN-BET interaction contributes to mitotic chromatin tethering of the MLV PIC is not known . We therefore infected cells with VLPs carrying a mutant IN that is deficient for BET-protein binding , IN_W390A [29] . WT p12 and CA puncta were still observed to co-localise with mitotic chromatin in these cells ( Fig 2C ) , suggesting that the IN-BET interactions are not essential for chromatin tethering during mitosis . Furthermore , whereas VLPs carrying mutations in the CTD of p12 ( p12_mut14 ) were severely defective in replication ( >150-fold compared to WT VLPs ) , IN_W390A containing VLPs showed only a mild defect of <2-fold in our assays ( Fig 2D ) . Our results therefore suggest that IN-BET interactions may not be as critical for MLV replication per se compared to p12-chromatin interactions . Interestingly , whereas the insertion of a heterologous chromatin binding sequence ( hCBS ) from the prototypic foamy virus ( PFV ) into p12 increased the replication of p12_mut14 carrying VLPs , it did not rescue the slight infectivity defect of the IN_W390A mutation ( Fig 2D ) . Thus , p12- and IN-chromatin interactions likely have distinct functions for the virus . Having shown that the chromatin localisation of CA-containing PICs in infected cells was dependent on the p12 CTD , we next wanted to investigate the interactions of the p12 CTD with cellular proteins . We therefore synthesised N-terminally GST-tagged Mo-MLV p12 and analysed its expression and sub-cellular localisation . Biochemical fractionation of cycling 293T cells transiently expressing GST-p12 revealed that , surprisingly , recombinant WT p12 was mainly cytoplasmic , being detected in the same fraction as HSP90 , a cytosolic marker , and not in the fraction containing histone H2B , a chromatin marker ( Fig 3A , lanes 1–3 ) . Similar localisation was observed with GST-tagged p12_mut 14 ( Fig 3A , lanes 4–6 ) . The PFV hCBS is known to bind histones H2A-H2B [14 , 15] and GST-p12 carrying this motif ( p12+hCBS ) could be readily detected in the chromatin pellet fraction ( Fig 3A , lane 9 ) , suggesting that the GST-tag itself is unlikely to be sterically interfering with chromatin tethering . In general , mitotic cells comprise only a small fraction of a cycling cell population . Therefore , to distinguish between p12 localisation in interphase and mitotic cells , we probed the sub-cellular distribution of p12 by immunostaining HeLa cells stably-expressing GST-p12 ( Fig 3B ) . In agreement with the fractionation assays , stably-expressed p12+hCBS was enriched in the nuclei of interphase cells and localised on the chromatin of mitotic cells ( Fig 3B ) . However , both p12_WT and p12_mut14 were mainly cytosolic in interphase cells and did not show a detectable association with mitotic chromatin in the few mitotic cells present ( Fig 3B ) . To enrich cells in mitosis and look directly for an interaction with chromatin , we arrested 293T cells transiently expressing GST-p12 ( ~38 kDa ) in metaphase , by nocodazole treatment , prior to their lysis for pull-down assays . First , cell lysates normalised for total protein concentration were incubated with glutathione-sepharose beads . The eluates from the beads were subsequently analysed by SDS-PAGE followed by silver-staining ( Fig 3C , left panel ) and immunoblotting ( Fig 3C , right panel ) . The core histones , H2A , H2B , H3 and H4 , were all pulled-down with p12+hCBS in our assays ( Fig 3C , lane 4 ) but not with p12_WT ( Fig 3C , lane 2 ) . These GST pull-down assays were performed with lysates pre-treated with nuclease to eliminate the detection of DNA- or RNA-mediated protein interactions . Therefore , to determine whether GST-p12 was able to bind DNA we also incubated mitotic lysates of transfected cells with DNA-coated cellulose beads . HA-tagged IN , used as a positive control , showed binding to beads coated with both single- and double-stranded DNA ( Fig 3D , bottom panel ) . However , no binding to DNA was observed with GST-p12 ( Fig 3D , middle panel ) . Some additional minor bands were seen on both GST-p12 and IN-HA blots , which were likely break down products . Overall , four separate assays suggested that GST-p12 is unable to bind chromatin when expressed recombinantly . This concurs with a previous study , in which a Mo-MLV p12 fusion protein carrying a C-terminal GFP tag did not show detectable association with mitotic chromatin by fluorescence microscopy [19] . In that study , the inability of p12-GFP to associate with chromatin was attributed to a lack of phosphorylation of recombinant p12 [19] . To test this hypothesis in our system , we determined the phosphorylation status of GST-p12 in mitotic cells . Firstly , GST-p12 pulled-down from mitotic cell lysates was analysed by SDS-PAGE followed by sequential staining with two fluorescent dyes , ProQ diamond ( PQ ) and Sypro Ruby ( SR ) [30 , 31] . PQ is a small molecule fluorophore that binds with high-specificity to phosphorylated proteins . SR binds all proteins and is useful as a total protein indicator . The ratio of PQ and SR dye signals provides a measure of the phosphorylation level with respect to the total amount protein . Viral p12 is known to be mainly phosphorylated at serine 61 within the CTD [19 , 25] . We therefore compared the relative phosphorylation levels of GST-p12_WT with a S61A mutant . Compared to p12_S61A , p12_WT was at least 3-fold more phosphorylated ( Fig 3E , lanes 2 and 3 ) . Treatment of p12_S61A with alkaline phosphatase further reduced the phosphorylation level by ~2-fold ( Fig 3E , lanes 1 and 2 ) , suggesting that recombinant p12 is also phosphorylated at sites other than S61 . We next used nanoscale liquid chromatography coupled to tandem mass spectrometry ( nano LC–MS/MS ) to identify phosphopeptides of GST-p12_WT precipitated from 293T cells . The phosphorylation sites in the high confidence tryptic-peptides ( <1% false discovery rate , FDR ) were assigned using the probability-based phosphoRS algorithm [32] ( Table 1 ) . A peptide was counted as being phosphorylated at a specific residue if its phosphoRS site probability was >50% . For each tryptic peptide , the phosphorylation level at a particular site was estimated by dividing the phosphorylated peptide count by the total peptide count . This analysis , carried out with two biological replicates , confirmed S61 to be the main site of phosphorylation on GST-p12_WT , with ~45% of S61-containing peptides phosphorylated at this residue ( Table 1 ) . We also identified T52 , S65 , S78 , T80 and S81 as additional sites of low-level phosphorylation on p12 ( Table 1 ) . Our results suggest that , like viral p12 , recombinant GST-p12 is phosphorylated at S61 . The absence of a detectable association between GST-p12 and mitotic chromatin is therefore unlikely to result from a lack of phosphorylation of recombinant p12 . The inability of recombinant Mo-MLV p12 to bind mitotic chromatin even when phosphorylated suggests that the affinity of p12 for chromatin may be influenced by other viral factors . We have observed viral p12 to be bound to CA when it associates with mitotic chromatin in Mo-MLV infected cells ( Fig 2A ) . Prior to Gag cleavage , p12 is thought to be in a largely unstructured conformation [33] . In mature virions , the interaction of the p12 NTD with the CA lattice may induce a conformational change that increases the affinity of the p12 CTD for chromatin . In fact , such a conformational switch could potentially be important in the temporal regulation of late and early life cycle events of p12 . As part of Gag , p12 is known to interact with homologous to E6AP COOH terminus ( HECT ) ubiquitin ligase , WWP2 , via the Late ( L ) -domain motif , and with clathrin , CLTC , via the N-terminal DLL motif [5 , 34] . As these interactions do not require p12 to be bound to CA , they should , in theory , be recapitulated by recombinant GST-p12 in our system . To test whether GST-p12 was acting like the p12 region of Gag , we used a global proteomic approach based on stable isotope labelling by amino acids in cell culture ( SILAC ) -mass spectrometry ( MS ) to look for host proteins that interact with GST-p12 [35 , 36] . A schematic diagram of the workflow is illustrated in Fig 4A . Briefly , GST , GST-p12_mut14 and GST-p12_WT were transiently-expressed in 293T cells cultured in light ( R0/K0 ) , medium ( R6/K4 ) and heavy ( R10/K8 ) SILAC media , respectively . Next , lysates of nocodazole treated transfected cells were prepared , normalised and used for parallel pull-down assays with glutathione-sepharose beads . The eluates from the beads were then pooled in a 1:1:1 ratio and subjected to LC-MS/MS analysis . The experiment was performed using biological replicates to test reproducibility of the mass-spec hits . The pull-down assays for this experiment were performed using mitotic cell lysates to facilitate identification of any potential chromatin interactions that were previously missed . After removal of obvious contaminants ( e . g . keratin ) and non-quantified hits , 948 and 879 proteins were identified in replicates 1 and 2 ( R1 and R2 ) of the experiment respectively ( the false discovery rate ( FDR ) was set at 5% ) . To select proteins enriched in the heavy-labelled ( H ) GST-p12_WT sample relative to the light-labelled ( L ) GST sample , we compared the log2 ( H/L ) silac ratios of the mass-spectrometry hits . When plotted as a frequency distribution , the log2 ( H/L ) ratios of each replicate fitted well to a normal distribution curve ( R2 ≥ 0 . 98 ) , allowing the mean and standard deviation to be accurately estimated ( Fig 4B ) . In this experiment , contaminants/non-specifically bound cellular proteins were assumed to cluster around the mean of the distribution . Therefore , hits with log2 ( H/L ) ratios greater than 2 . 58 standard deviations ( SDs ) from the mean were considered to be significantly enriched ( 99% confidence threshold , p≤0 . 01 ) in the GST-p12_WT sample . Twenty one such proteins were identified in R1 and 22 identified in R2 ( Fig 4B Venn diagram ) . Of these , seven proteins were found in both R1 and R2 , suggesting that they may bind recombinant p12_WT ( Fig 4B and S1 Table ) . These proteins were then ranked based on normalised abundance in the group . In mass-spectrometry analysis , larger proteins are in general assigned higher scores . Therefore , to control for protein size , we divided the number of peptide spectral matches identified for each protein with its length and then normalised the values across the group ( Fig 4C and 4D ) . When ranked this way , the clathrin heavy chain ( CLTC ) , clathrin interactor 1 ( CLINT1 ) and NEDD4-like E3 ubiquitin ligase ( WWP2 ) were identified as the top three hits . As CLTC and WWP2 are known interactors of the p12 region of Gag , our results suggest that recombinant p12 may be more representative of immature p12 . The function of clathrin in MLV infection is unclear . Interestingly , in mitosis , CLTC has recently been found to play a role in cross-linking the kinetochore microtubules of the spindle [37] . Furthermore , of the seven proteins identified as p12_WT binders in this experiment , four ( including CLINT1 ) have previously been found to associate with CLTC in a mitotic spindle stabilising complex [38] . Thus , clathrin binding may be relevant to the early stages of MLV replication and should be investigated further . To validate these mass-spectrometry results , we next performed glutathione-sepharose pull-down assays with mitotic lysates containing GST-tagged WT p12 and a panel of p12 mutants ( the mutations are annotated in Fig 1A ) . The eluates from the assays were subsequently analysed by silver-staining ( Fig 4E , left panel ) and immunoblotting ( Fig 4E , right panel ) . CLTC and CLINT1 were pulled-down with all p12 proteins apart from the NTD mutants 7 and 8 ( Fig 4E , lanes 4 and 5 ) . This is consistent with the location of the CLTC-binding motif , DLL in p12: In p12_mut8 , the DLL motif is substituted with alanines . In p12_mut7 , residues neighbouring the DLL motif are mutated which may also inhibit CLTC recruitment . The interaction between p12 and WWP2 was also re-capitulated as expected in the assays , with WWP2 being pulled-down with all p12 proteins apart from the L-domain mutant and p12_mut8 . To identify potential chromatin interactions of p12 , we compared the heavy-labelled ( H ) GST-p12_WT sample to the medium-labelled ( M ) GST-p12_mut14 sample , as viruses carrying p12_mut14 do not localise to chromatin ( Fig 2B ) . However , after performing a similar analysis with the log2 ( H/M ) silac ratios of the mass-spectrometry hits as above ( S2 Fig ) , we found no proteins that passed the criteria for significant enrichment ( log2 ( H/M ) ratio > 2 . 58 SDs from the mean ) in both replicates ( Fig 4D and S2 Fig ) . The lack of a significant difference in the interactions of recombinant p12_WT and p12_mut14 suggests that the CTD functions of viral p12 are not replicated by GST-p12 in this system . Based on the observations above , recombinant GST-p12 appears to mimic Gag-p12 as it was able to recapitulate the known interactions of the p12 region of Gag , which are mediated by motifs within or next to the NTD region of p12 . However , GST-p12 did not show detectable chromatin interactions , unlike viral p12 in the PIC . We therefore suggest that another viral protein , probably CA , influences the ability of p12 to interact with chromatin . There are a number of potential mechanisms by which CA may facilitate the interaction of the p12 CTD with mitotic chromatin . If the p12 NTD sterically-hinders the CTD from interacting with chromatin , the binding of CA to the NTD could remove this inhibition . On the other hand , CA binding may prevent host proteins , like clathrin , from binding to p12 and sterically hindering a chromatin interaction . Alternatively , binding to CA may induce a subtle but important conformational change in the peptide backbone of p12 which increases its affinity for chromatin . To potentially distinguish between these possible mechanisms , we introduced a range of alterations into Mo-MLV GST-p12 and tested the mutants for chromatin interactions . In the p12_CTD only mutant , the p12 NTD was deleted to remove any steric hindrance from this domain . In the p12_D25A/L-dom mutant , the CLTC and WWP2 binding sites were removed by substituting alanine at D25 of the DLL motif and at the PPPY L-domain motif , respectively . The other two mutants included in the panel , p12_M63I and p12_G49R/E50K , carried changes that were previously identified in an in vitro evolution study to find viral revertants that compensated for the non-infectious Mo-MLV p12_SS ( 61 , 65 ) AA mutant that lacks phosphorylation [25] . Recently the M63I substitution has also been shown to rescue chromatin tethering of recombinant Mo-MLV p12-GFP [19] . We first tested the panel of p12 mutants in pull-down assays from mitotic cells using glutathione-sepharose beads ( Fig 5A ) . GST-tagged p12_WT and p12+hCBS were included in these assays as negative and positive controls for chromatin binding , respectively . Analysis of bead eluates by silver-staining and western blotting revealed that core histones were pulled-down with p12_M63I ( Fig 5A , lane 2 ) but not with the other mutants ( Fig 5A , lanes 3–5 ) . Probing the sub-cellular distribution of the mutants by immuno-staining of stably-transduced HeLa cell lines ( S3 Fig ) only detected a chromatin association of p12_M63I and p12+hCBS during mitosis . In agreement with the immunoprecipitation data , the other mutants were excluded from mitotic chromatin . Thus , removing the p12 NTD or preventing known interactions with host proteins at/near the NTD was unable to confer chromatin binding to GST-p12 . Instead , this was attained by the M63I substitution in the p12 CTD . Surprisingly , the G49R/E50K mutation , which also rescues the infectivity of the SS ( 61 , 65 ) AA mutant , was unable to confer chromatin tethering of GST-p12 in our assays . M63I , is located within the p12 CTD , whereas the G49R/E50K changes are located outside of the CTD region . Therefore , perhaps only the M63I change alters the conformation of the chromatin binding region of p12 directly . Overall , our results suggest that the inability to detect an interaction between phosphorylated GST-p12_WT and chromatin in the absence of CA is probably not due to steric inhibition by the p12 NTD or cellular proteins but due to the p12 CTD being in a conformation that is not conducive for detectable binding . We also tested the M63I and G49R/E50K mutants in infectivity assays ( Fig 5B ) . VLPs carrying p12_G49R/E50K showed similar infectivity to WT Mo-MLV , whereas p12_M63I VLPs had a mild ~2-fold defect . As p12+hCBS VLPs also show a moderate defect ( ~5-fold ) in infectivity compared to WT Mo-MLV , our results corroborate the hypothesis that increasing the affinity of Mo-MLV p12 for chromatin , above WT levels , may be detrimental to viral replication [12] . Furthermore , the M63I and G49R/E50K mutations were unable to rescue the infectivity of p12_mut14 VLPs ( Fig 5B ) . This implies that the M63I mutation modulates an existing p12 interaction rather than providing an independent interaction like the PFV hCBS . It was somewhat surprising that the conservative single amino acid change of methionine-63 to isoleucine could have a dramatic effect on the chromatin interaction of p12 . Interestingly , we noted that the p12 orthologues from N-tropic ( N ) -MLV and feline leukaemia virus ( FeLV ) naturally had an isoleucine residue at position 63 ( Fig 5C ) . We have previously shown that other gammaretroviral p12 proteins have similar functions to Mo-MLV p12 [7] and so wondered whether these orthologues may interact with chromatin when expressed recombinantly . We therefore expressed p12 from different gammaretroviruses as GST fusion proteins and tested them in pull-down assays ( Fig 5D ) . Excitingly , core histones were pulled-down with p12 proteins from FeLV , gibbon ape leukaemia virus ( GaLV ) and koala retrovirus ( KoRV ) ( Fig 5D , lanes 4 , 7 and 8 ) , but not with p12 from xenotropic MLV-related virus ( XMRV ) or N-MLV ( Fig 5D , lanes 5 and 6 ) . Recombinant gammaretroviral p12 proteins therefore appear to differ in their interactions with chromatin , with some p12-chromatin interactions being detectable in pull-down assays in the absence of CA and other viral proteins . The p12 CTD regions of FeLV , N-MLV and XMRV are >68% identical to Mo-MLV in sequence . The amino acids at residues equivalent to 63/64 in the different viruses are: M/A in Mo-MLV , I/A in FeLV , I/V in N-MLV and M/V in XMRV . As changing M63 to I , making it I/A , enhanced the chromatin interaction of Mo-MLV p12 , and FeLV also has I/A at these positions and interacts with chromatin , we decided to investigate these residues further . We tested the effects of I52M and A53V mutants of FeLV p12 in pull-down assays . However , both I52M and A53V showed similar levels of histone precipitation ( <2-fold difference ) as the WT protein ( Fig 5E ) , implying that the extent to which the I/A motif influences chromatin binding may be dependent on its sequence-context . Indeed , GaLV and KoRV p12 proteins which show only ~27% identity to Mo-MLV p12 in the CTD appear to interact with chromatin with sufficient affinity for in vitro detection in the absence of this motif . As the mode of chromatin binding appears to be conserved between Mo-MLV p12_WT and p12_M63I , both being dependent on residues 65–69 ( Mut14 ) in p12 , we used recombinant p12_M63I , which showed an association with mitotic chromatin in our biochemical assays , to characterise this interaction further . In virus-infected cells , Mo-MLV p12 has only been observed to associate with chromatin between nuclear envelope disassembly and chromatin decondensation ( Fig 2A ) [11] . However , whether this reflects a preferential affinity of p12 for mitotic chromatin or merely the accessibility of chromatin during infection is unclear . Pre-mitosis , the MLV PIC is unable to actively traverse through the nuclear pores and post-mitosis , the dissociation of p12 from chromatin could be driven by other viral factors in the PIC . To investigate the cell-cycle dependency of the p12-chromatin interaction directly , we compared the chromatin interaction of recombinant Mo-MLV p12_M63I in interphase and mitotic cells . We first analysed the sub-cellular localisation of transiently-expressed GST-p12_M63I in cycling 293T cells using biochemical fractionation . When probed by immunoblotting , very little p12_M63I was detected in the chromatin pellet fraction ( Fig 6A , lane 3 ) compared to the p12+hCBS control ( Fig 6A , lane 6 ) . To distinguish between p12_M63I localisation in interphase and mitotic cells , we next investigated the sub-cellular distribution of stably-expressed GST-p12_M63I by immunofluorescence . Whereas , GST-p12+hCBS was enriched on chromatin in both interphase and mitotic HeLa cells ( Fig 6B , bottom panels , red and blue box , respectively ) , GST-p12_M63I showed a differential localisation ( Fig 6B , top panels ) . It was chromatin-associated in mitosis ( blue box ) , but mainly cytosolic in interphase cells ( red box ) . The difference in the distribution of GST-p12_M63I and GST-p12+hCBS in interphase is likely due to differences in chromatin affinity rather than nuclear access . The hCBS sequence does not include a nuclear localisation signal [39] and as both GST_p12_M63I and GST_p12+hCBS are of similar size , they would be expected to progress at similar rates across the nuclear envelope by passive diffusion . As GST-p12_M63I appeared to have a higher affinity for chromatin in mitosis compared to interphase , we performed glutathione-sepharose bead pull-down assays from cells arrested primarily in mitosis ( >80% of cells in G2/M ) or interphase ( <2% of cells in G2/M ) following treatment with nocodazole or aphidicolin , respectively . The final eluates from the beads were analysed by silver-staining and immunoblotting ( Fig 6C , top and bottom panels , respectively ) . Approximately 4-fold more histones were pulled down with GST-p12_M63I from mitotic cell lysates than interphase cell lysates ( Fig 6C , lanes 2 vs 5 , Fig 6D ) , whereas GST-p12+hCBS showed similar levels of histone precipitation from both interphase and mitotic cell lysates ( Fig 6C , lanes 3 vs 6 , Fig 6D ) . The higher affinity of GST-p12_M63I for chromatin in mitosis could be due to changes in the post-translational state of p12 itself , increased expression of specific interacting proteins or changes in the putative chromatin target during this stage of the cell cycle . As phosphorylation of many proteins is cell-cycle dependent , we investigated whether the phosphorylation state of p12 changed in mitosis . For this purpose , GST-p12 proteins precipitated from mitotic ( nocodazole-treated ) and interphase ( aphidicolin-treated ) cell lysates were analysed by SDS-PAGE followed by sequential staining with PQ and SR dyes . From the PQ/SR ratios , GST-p12_WT and GST-p12_M63I were observed to be at least ~3–4 fold more phosphorylated at mitosis in comparison to interphase ( Fig 6E , lanes 1 and 2 vs 3 and 4 ) , which correlated well with the increased chromatin interaction observed for p12_M63I during mitosis . The S61A mutant was included as a negative control for phosphorylation . As both chromatin interaction and phosphorylation of GST-p12_M63I appear to increase in mitosis , we next investigated whether phosphorylation directly influences the chromatin interaction of p12 by exploring the effects of kinase inhibition and phospho-ablative/-mimetic mutations on histone precipitation . Glycogen synthase kinase 3 ( GSK3 ) and cyclin-dependent kinase 5 ( CDK5 ) were predicted to phosphorylate Mo-MLV p12 on S61 , by the NetPhos 3 . 1 kinase prediction algorithm [40 , 41] . LiCl and roscovitine have been shown to inhibit GSK3 and CDK5 respectively , whereas kenpaullone inhibits both enzymes [42] . These drugs were tested in our biochemical assays for their ability to reduce GST-p12_M63I phosphorylation and chromatin association . In our assays , cells transiently-expressing GST-p12_M63I were cultured in media containing nocodazole overnight , prior to the addition of the kinase inhibitors . Kinases were only inhibited for a short period of time ( 3 . 5 h ) in the presence of both nocodazole and MG132 ( proteasome inhibitor ) , to avoid both cell lethality and exit from mitosis . After removal of the drugs , cells were immediately washed and lysed for glutathione-sepharose bead pull-down assays . The eluates from the beads were analysed by SDS-PAGE followed by sequential staining with PQ and SR dyes to quantify the phosphorylation levels of GST-p12_M63I . LiCl treatment had little effect on GST-p12_M63I phosphorylation ( Fig 7A , lane 2 ) . Roscovitine and kenpaullone reduced phosphorylation by ~1 . 5-fold and ~3-fold respectively ( Fig 7A , lanes 3 and 4 ) . To correlate GST-p12_M63I phosphorylation with its ability to interact with chromatin , we quantified the amounts of H2B in the pull-down eluates by immunoblotting . Whereas LiCl treatment had no significant effect on chromatin pull down ( Fig 7B , lane 2 ) , roscovitine and kenpaullone reduced H2B precipitation by ~1 . 5-fold and ~3 . 5-fold , respectively ( Fig 7B , lanes 3 and 4 ) , reflecting the observed decrease in phosphorylation . The short period of kinase inhibition used in our assays , may have limited the effects observed on GST-p12_M63I phosphorylation . However , our results suggest that reduction of p12 phosphorylation decreases its affinity for chromatin . To explore the relationship between p12 phosphorylation and chromatin interaction further , we made phosphoablative ( S61A ) and ‘phosphomimetic’ ( S61D and S61E ) mutations in GST-p12_M63I . When tested in glutathione-sepharose bead precipitation assays , the S61A mutant showed an ~3-fold reduction in histone pull down ( Fig 7C , lane 2 ) . The decrease in chromatin association due to the loss of phosphorylation was partly rescued in the S61E mutant ( ~1 . 5-fold defect ) but not in S61D ( ~6-fold defect ) ( Fig 7C , lanes 3 and 4 ) . The extent to which aspartic acid and glutamic acid mimic the effect phosphorylation is heavily context-dependent , as these amino acids differ from phosphate in charge ( -1 versus -1 . 5 ) and the number of oxygen atoms available for hydrogen bonding . In the context of the p12 protein , S61E might be better tolerated than S61D , as glutamic acid is more similar in size and geometry to phosphorylated serine than aspartic acid [43] . We also tested the effects of these phosphoablative and ‘phosphomimetic’ mutations on Mo-MLV infectivity . In the p12_M63I background , all three mutations increased infectivity slightly , by ~2-fold or less ( Fig 7D ) . This was not surprising as the M63I mutation was initially described to be a compensatory mutation for loss of phosphorylation . However , in p12_WT , all S61 mutations decreased infectivity by >40-fold ( Fig 7D ) , suggesting that S61 may make important contributions to the chromatin interaction beyond phosphorylation . GST-p12_M63I was seen to colocalise with mitotic chromatin and to precipitate histones in GST-pull down assays ( Figs 5 and 6 ) . However , it could be interacting directly with components of chromatin or indirectly via other chromatin-binding proteins . To potentially identify the target of p12_M63I , we compared the interactome of GST-p12_M63I ( which showed chromatin association ) to that of GST-p12_WT ( which did not ) in 293T cells , using a SILAC-MS approach ( Fig 8 ) . Light ( R0/K0 ) and medium ( R6/K4 ) SILAC-labelled cells transiently expressing GST-p12_M63I and GST-p12_WT respectively were treated with nocodazole for mitotic enrichment and then lysed for glutathione-sepharose bead pull-down assays . Eluates were subsequently pooled in a 1:1 ratio for analysis by LC-MS/MS . As before , the experiment was performed using two biological replicates to test reproducibility . The mass-spec hits identified at 5% FDR were subjected to further downstream analysis , in order to select proteins significantly enriched in the light-labelled ( L ) GST-p12_M63I sample relative to the medium-labelled ( M ) GST-p12_WT sample . For 68 of these proteins the log2 ( L/M ) ratios were greater than 2 . 58 standard deviations ( SDs ) from the mean ( 99% confidence threshold ) in both replicates , suggesting that they may interact with GST-p12_M63I ( Fig 8A , S2 Table ) . Interestingly , InterPro domain analysis [44] of these proteins revealed the vast majority to be involved in the regulation of chromatin structure and function ( Fig 8B ) . Furthermore , when ranked on normalised abundance within the group , the highest-scoring hits in both replicates were nucleosomal core histones . The other top-ranking hits , with a normalised abundance ≥1 in both replicates , included chromobox homolog proteins and components of the chromosomal passenger complex ( Fig 8C ) . To validate the mass-spectrometry results , we next performed pull-down assays with mitotic cell lysates containing GST-tagged p12_WT , p12_mut14 , p12_M63I , p12_M63I/mut14 , p12+hCBS and p12+hCBS/mut14 proteins . The eluates from the assays were analysed by silver-staining and immunoblotting for the top protein hits for GST-p12_M63I in the mas-spec analysis ( Fig 8D ) . All proteins tested were precipitated by p12+hCBS and p12+hCBS/mut14 as well as p12_M63I ( Fig 8D , lanes 3 , 5 and 6 ) . However , none were precipitated with the negative controls p12_mut14 and p12_M63I/mut14 ( Fig 8D , lanes 2 and 4 ) , or the p12_WT , protein ( lane 1 ) . As p12_M63I and PFV hCBS showed similar chromatin interactions ( Fig 8D ) , we performed another SILAC-MS experiment to compare their global interactomes . GST-p12+hCBS/mut14 was used to ensure that all of the observed interactions were specific to the hCBS and that there was no contribution from the CTD of p12 . Both GST-p12_M63I and GST-p12+hCBS/mut14 pull-down eluates were compared to GST-p12_M63I/mut14 pull-down eluates , to identify their putative chromatin interactions . Of the 73 proteins identified as GST-p12_M63I interactors from this analysis , ~90% ( including 39 of the 40 top-ranked hits ) overlapped with the GST-p12+hCBS/mut14 interactome ( S4 Fig ) . Together with the observation that the top SILAC-MS hits for p12_M63I were core histones , this suggests that p12_M63I may also bind chromatin directly like the PFV hCBS [14 , 15] . The PFV hCBS is known to bind to the nucleosomal core histones H2A and H2B [14 , 15] . To identify whether p12 proteins also interact directly with nucleosomes , we used biolayer interferometry ( BLI ) to probe the binding of recombinant poly-nucleosomal arrays by p12 CTD peptides ( Fig 9 ) . Streptavidin sensor probes coated with biotinylated peptide ligands were immersed in ‘analyte’ solutions containing the in vitro assembled nucleosomal arrays to observe any potential interactions . The p12 CTD peptide sequences used are highlighted in bold in Fig 5C . A peptide carrying the PFV hCBS sequence was included in the assays as a positive control . We first tested the binding of Mo-MLV p12_M63I CTD peptides to nucleosomes ( Fig 9A ) . Excitingly , both S61-phosphorylated and non-phosphorylated p12_M63I peptides showed clear binding to nucleosomes ( Fig 9A , orange and red lines ) . This binding was specific as an interaction was not observed with the negative control p12_M63I_R66A peptide ( Fig 9A , dark red line ) . R66 is part of the ‘SRLRG’ motif in the p12 CTD which is required for tethering to mitotic chromatin ( Fig 2B and Fig 8D ) and we have previously shown that Mo-MLV VLPs carrying p12_R66A are non-infectious [7] . To compare the affinities of PFV hCBS and phosphorylated/non-phosphorylated p12_M63I for nucleosomes , the peptides were also tested against a dilution series of the nucleosomal preparation ( Fig 9B ) . The equilibrium dissociation constant ( Kd ) measurements derived from this experiment were in the low nanomolar range ( ~2–3 nM ) for all peptides , indicating that the peptides bound very strongly to the nucleosomes . Furthermore , the estimated Kd measurements for hCBS , phosphorylated p12_M63I and non-phosphorylated p12-M63I peptides in this experiment were very similar ( <2-fold variance ) . Although somewhat surprising , the low Kd values may arise from the high avidity of the polymeric nucleosome arrays ( up to 11 nucleosomes per DNA molecule ) used . Compellingly , p12 CTD peptides from FeLV and GaLV , two gammaretroviruses that also demonstrated chromatin interactions in our GST pull-down assay ( Fig 5D ) , also showed high affinity ( nM ) binding to recombinant nucleosomal arrays ( Fig 9C ) . Although we could not detect any interactions of GST-p12_WT with chromatin in our previous assays , we reasoned that the high sensitivity of BLI coupled with the high avidity of our polymeric nucleosome arrays may sufficiently compensate for a potential low affinity interaction between Mo-MLV p12_WT and chromatin allowing us to detect binding . Indeed , provokingly , a biotinylated peptide corresponding to the CTD sequence of Mo-MLV p12_WT showed measurable binding to nucleosomal arrays by BLI ( Fig 9D ) . This suggests that WT p12 does bind directly to chromatin in vitro but that the affinity is too low to detect the interaction in most standard assays . To identify whether p12 binds to the same nucleosomal docking site as the CBS from PFV Gag , we then tested the ability of p12 CTD peptides to block the binding of nucleosomal arrays to hCBS and vice versa ( Fig 9E and 9F ) . In the first set of these experiments , nucleosomal arrays were pre-incubated with excess non-biotinylated Mo-MLV p12_WT , Mo-MLV p12_R66A , hCBS or hCBS_L539A/R540A peptides , prior to recording BLI measurements with biotinylated hCBS immobilised on sensor probes ( Fig 9E ) . L539A/R540A mutations have previously been observed to prevent binding of hCBS peptides to nucleosomes [15] . Equilibrium binding of nucleosomes to immobilised hCBS was significantly decreased in the presence of both soluble hCBS ( by ~10-fold ) and p12_WT ( ~4-fold ) , but not hCBS_L539A/R540A or p12_R66A peptides ( Fig 9E ) . In reciprocal experiments , a soluble non-biotinylated hCBS peptide could compete with the binding of immobilised Mo-MLV p12_WT , Mo-MLV p12_M63I , FeLV and GaLV to the nucleosomal arrays ( Fig 9F ) . Overall , these results suggest a conserved mode of chromatin binding between spuma and gamma retroviruses .
It has long been known that the Gag-cleavage product p12 is essential for MLV replication [6] . As well as the late-domain required for budding , p12 contains N- and C-terminal functional domains that are required for the early stages of replication [6 , 7] . We previously showed that p12 binds CA and that mutating the NTD of p12 results in reduced stability and abnormal morphology of viral cores leading to reduced infectivity [9] . Additionally , mutating the CTD of p12 prevents viral PICs from associating with chromatin and reduces infectivity to less than 1% . The infectivity of p12 CTD mutants can be partially restored by inserting a heterologous CBS into p12 [7 , 11 , 12] . Here , we confirm and extend these observations and provide further mechanistic insights into the functions of both N- and particularly C-terminal p12 domains . In this study , we showed that the NTD of p12 interacts specifically with a regular hexameric arrangement of CA and not individual CA monomers in vitro ( Fig 1 ) . Moreover , we were also able to detect an association of p12 and CA in mature viral particles by co-immunoprecipitation . Mutations to the CTD of p12 did not disrupt this interaction , either in viruses or in in vitro binding assays , suggesting that the CTD of p12 does not make CA interactions , and that any potential NTD-CTD interactions within p12 are not required for CA binding . Importantly , this also implies that CA binding does not modify the CTD of p12 directly . Specific binding to mature CA lattices supports previous findings that immature particle assembly is not affected by p12 mutations [9] , and that in a NMR study of a Gag-like fragment , no long-range interactions could be detected between p12 and the N-terminal region of CA [33] . Intriguingly , we calculated a p12:CA binding ratio of 1:6 from our in vitro binding experiments ( S1 Fig ) that may be indicative of p12 binding to hexameric rings of CA monomers . However , this may be an under-estimate , as p12 may not occupy all available sites on the lattice , even when present in molar excess , and it is possible that not all the CA on the lipid tubes will form regular hexamers [26] . Further studies will be required to identify the p12 binding site on CA , which , in turn , may reveal how p12 stabilises the CA shell . Interestingly , we observed co-localisation of CA with p12 in infected cells ( Fig 2 ) , even on mitotic chromatin . Notably , our findings differ from a previous study in which co-localisation of CA with p12 was observed in the cytosol of infected cells but not on mitotic chromatin [11] . However , as the antibodies and fixative procedures used in that study are different from the ones used here , we believe that the discrepancy in the observations could be due to different sensitivities of CA detection . The CA shell of HIV-1 is proposed to disassemble before integration . Although the timing of such “uncoating” is controversial , it appears to be triggered after the first strand transfer step of reverse transcription [45] . Little is known about the uncoating of other retroviruses , but MLV CA has also been observed to dissociate from the PIC gradually with time [9] . This implies that the amount of CA still associated with the PIC upon chromatin tethering is likely to be less than in the viral core , making it harder to detect . We previously found that the fixative procedures for the anti-CA and anti-p12 antibodies were not compatible , and we therefore stained for p12 and CA on separate slides [9] . Here , we introduced a two-step fixation procedure ( 4% PFA and methanol ) and optimised membrane permeablisation of our samples , which now allows us to robustly detect both p12 and CA simultaneously . When we introduced mutations into the CTD of p12 , the protein remained co-localised with CA but no longer associated with chromatin . In contrast , mutating IN to prevent interaction with BET proteins had no effect on CA or p12 localisation . This implies that CA association with mitotic chromatin is primarily driven by p12 . MLV CA is a component of the PIC [46] but whether it plays a role in the chromatin targeting of the viral genome like HIV CA [47] is currently not known . Recently the interaction of MLV IN with BET proteins was shown to mediate integration site selection in infected cells [20–22] . We showed here that VLPs carrying an IN mutant deficient in BET binding were only mildly defective ( <2-fold ) in infectivity compared to WT virus , whereas p12 CTD mutant VLPs were less than 1% infectious . These results suggest that p12 targeting to mitotic chromatin is mandatory for integration , and likely occurs prior to IN-chromatin binding . Enticingly , this is reminiscent of the proposed interplay between the HIV-1 CA-binding host protein , cleavage and polyadenylation specificity factor 6 ( CPSF6 ) and IN-binding lens epithelium-derived growth factor ( LEDGF/p75 ) interactions in the nuclear targeting of HIV-1 [47] . CPSF6 is currently thought to direct the HIV-1 PIC to transcriptionally-active chromatin , and then LEDGF/p75 mediates the local targeting of integration specifically into genes . MLV p12 could therefore play a similar role to CPSF6 in facilitating nuclear/chromatin delivery of CA-containing viral PICs . This analogy could even be partly extended to foamy viruses that have a CBS within their Gag shell . This would suggest an evolutionary conserved overall pathway for retroviral chromatin targeting , albeit with different mechanistic details for individual viruses . It will be interesting to investigate the timing of p12 and CA dissociation from each other , and from chromatin , to determine the order of events relative to integration itself . Excitingly , we have shown for the first time that WT Mo-MLV p12 CTD , but not the defective R66A CTD mutant , binds directly to nucleosomal arrays in vitro ( Fig 9 ) . This shows that p12 tethers the PIC to chromatin via a direct interaction and not through another chromatin binding protein . Furthermore , it confirms that only the CTD motif is necessary for the binding . Nucleosome binding was also observed with p12 orthologues of FeLV and GaLV , suggesting that direct chromatin tethering is a common feature of gammaretroviruses . Importantly , p12 CTD peptides were able to compete with the binding of the PFV CBS to nucleosomal arrays , and vice versa , in BLI assays , indicating that they all bind around the same site . The similarity between gammaretroviral p12 and PFV CBS chromatin binding is supported by our data from cell-based assays . Global mass-spectrometric analysis of the chromatin interactome of p12_M63I revealed a very significant overlap of ~90% with that of the PFV CBS ( Fig 8 and S4 Fig ) . The docking site of the PFV CBS on nucleosomal surfaces has been mapped by x-ray crystallography to an acidic patch at the H2A-H2B heterodimeric interface [15] . Interestingly , this is also the target site of the KSHV LANA protein on nucleosomes [18] . Both PFV CBS and KSHV LANA carry an arginine side chain that projects deep into the nucleosomal acidic pocket and makes critical contacts with its carboxylate groups . Gammaretroviral p12 proteins also carry a number of conserved arginine residues in the CTD which have been shown to be essential for infection and chromatin association ( Figs 2 , 8 and 9 ) [7] , suggesting that p12 proteins bind the same nucleosome docking site using a conserved mechanism . Our in vitro nucleosome binding assays recapitulated the mitotic chromosome tethering of Mo-MLV p12 seen in infected cells ( Fig 2 ) [11] . However , in order to characterise this interaction better in cells , we initially studied GST-tagged p12 proteins . Although we showed that GST-tagged FeLV , GaLV and the Mo-MLV p12_M63I mutant could precipitate histones from cell lysates ( Fig 5 ) , surprisingly , we could not detect a chromatin association in cells with GST-tagged WT Mo-MLV p12 ( Fig 3 ) or indeed other MLVs ( N-tropic MLV and XMRV , Fig 5 ) . There are a few possible explanations for this: A previous study attributed the inability of a WT Mo-MLV p12-GFP fusion protein to bind chromatin to an absence of phosphorylation [19] . However , ~50% of the GST-tagged Mo-MLV p12_WT protein expressed in our cells was clearly phosphorylated on S61 ( Fig 3 , Table 1 ) . The sensitivity of our assays would have allowed us to detect an interaction between p12 and chromatin in this fraction of p12 , indicating that the absence of detectable chromatin binding here is not due to a lack of phosphorylation . However , phosphorylation may be involved in modulating the affinity of p12 for chromatin , as reducing phosphorylation of GST-p12_M63I using kinase inhibitors or an S61A mutation correlated with reduced chromatin binding ( Fig 7 ) . Moreover , p12_M63I had a higher affinity for chromatin in mitosis compared to interphase ( Fig 6 ) [11] , and this correlated with an increase in phosphorylation during mitosis . Surprisingly , the phospho-mimetic mutations S61D and S61E also reduced chromatin binding . However , changing these residues may affect the hydrogen bonding capacity or conformation of p12 as well as phosphorylation . Notably , a link between serine/threonine phosphorylation and chromatin binding has been established for the LANA protein of KSHV . Preventing LANA phosphorylation by short- term treatment ( 4 h ) with a RSK inhibitor decreased H2B binding by ~50% [48] . Kinases which phosphorylate LANA also appear to phosphorylate the EBNA-1 protein of EBV , therefore chromatin binding of the latter may also be subjected to similar post-translational regulation [48] . Whether phosphorylation of LANA and EBNA-1 proteins is cell-cycle dependent is not known . As the inability of Mo-MLV GST-p12_WT to bind chromatin was not due to the lack of its phosphorylation , we wondered whether the absence of other viral proteins , particularly CA , may be steering GST-p12 to behave like the p12 region of Gag . In Gag , p12 recruits the ESCRT proteins required for viral budding through the L-domain PPPY motif and interacts with the clathrin heavy chain ( CLTC ) via the DLL motif in the NTD . As well as promoting budding , binding of these cellular proteins may disfavour chromatin binding . When a virus enters a new target cell , it must travel towards the nucleus , and at this time it would not be advantageous to recruit these membrane-associated host factors . Binding of the p12 NTD to CA may prevent interactions with these host proteins and instead facilitate the early steps of infection . Investigating the cellular interactome of Mo-MLV GST-p12_WT revealed that it did indeed bind WWP2 and CLTC ( Fig 4 ) . However , engineering p12 to remove the PPPY or DLL motifs , or the whole NTD , did not rescue chromatin tethering of GST-p12 ( Fig 5 and S3 Fig ) . Thus , chromatin binding is not significantly facilitated by preventing other host proteins from interacting with p12 . However , CA could increase the affinity of p12 for chromatin in alternative ways , perhaps by altering the confirmation of p12 , or by providing supplementary interactions to increase the affinity of binding . In our assays , chromatin binding of Mo-MLV GST-p12_WT was rescued by the M63I substitution in the p12 CTD . As GST-p12 is phosphorylated , M63I must be doing more than just compensating for a lack of this modification in our system . Importantly , the chromatin binding of this mutant was still dependent on the residues altered in Mut14 ( Fig 5 ) , suggesting that the M63I change is enhancing the ability of p12 to interact with chromatin without altering its mode of binding . M63I , therefore , probably increases the affinity of p12 for chromatin via a conformational change . This change may , at least in part , mimic the effect of CA-binding in the context of the viral PIC . The ability of M63I to rescue both phosphorylation-deficiency [25] and an absence of CA binding in the presence of phosphorylation ( Fig 6 ) , suggests a possible functional overlap between these two processes in enhancing chromatin binding of p12 . Another way in which CA could facilitate the chromatin tethering of the p12 CTD is by simply elevating the local concentration of p12 to increase the avidity of its interactions . The presence of multiple nucleosomes on the polymeric arrays used in our in vitro assay ( up to 11 per DNA ) may also increase the avidity of binding , allowing us to observe an interaction between WT Mo-MLV p12 that we did not observe in cells . A high avidity of polysomes could also explain why the different gammaretroviral p12 peptides had similar Kd values to each other and the CBS peptide in our BLI assays ( Fig 9 ) , despite the CBS showing greater histone binding in pull-down assays ( Fig 5 ) . Based on the results from this study and previous observations , we propose a model for the chromatin binding function of gammaretroviral p12 as shown in Fig 10A . As part of Gag , p12 exists in a largely unstructured conformation with high affinity for host proteins which facilitate viral budding but relatively low affinity for nucleosomes . Following Gag cleavage , p12 binds to the hexameric CA lattice via its NTD , stabilising the viral core and causing a change in the conformation of p12 which increases the affinity of p12 for nucleosomes . Upon breakdown of the nuclear envelope in mitosis , the PIC is targeted to condensed chromatin by CA-bound , phosphorylated p12 . Following exit from mitosis , either dephosphorylation of p12 or chromatin decondensation could reduce the affinity or avidity of the interaction and cause p12 to be released from chromatin . This may also release CA from the PIC and expose the viral cDNA and IN to chromatin . BET proteins then bind IN and direct the intasome to promotor regions where integration occurs . Failure of p12 to interact with chromatin leads to a severe replication defect ( <1% infectivity , Fig 2 ) . This may be partly due to exclusion from the re-assembled nucleus of unbound PICs , but likely indicates that p12-chromatin binding is required for an essential step preceding integration , like CA uncoating . Interestingly , in agreement with previous findings by Schneider et al . [12] , we have observed that alterations to the Mo-MLV p12 protein that increase the affinity for chromatin to a level that enables the interaction to be detected in microscopy or pull-down assays have a modestly detrimental effect on infectivity ( Fig 5 ) . These include the M63I substitution and the insertion of the PFV hCBS . Therefore , increased binding of p12 to chromatin may result in integration at the initial site of chromatin contact and prevent BET proteins targeting IN to optimal sites for integration . These results suggest that chromatin binding of gammaretroviral p12 is fine-tuned for optimal affinity and that stronger or weaker binding both appear to incur a fitness cost ( Fig 10B ) . In conclusion , we have demonstrated that gammaretroviral p12 docks to mitotic chromatin via a direct interaction with nucleosomal histones , similarly to spumaviruses . The chromatin binding of p12 is likely facilitated by binding to CA after viral maturation and by phosphorylation of p12 in mitosis . By tethering the CA-containing PIC to mitotic chromatin , p12 may influence global nuclear targeting of MLV integration , similarly to the proposed role of the CA-binding host protein CPSF6 in HIV-1 infection . Many retroviral genera have additional Gag-cleavage products in a similar genomic position to p12 , mostly with unknown functions but often containing a Late-domain . It is possible that they too have chromatin binding sequences and guide their viral PICs to chromatin . The efficiency of integration and of proviral transcription varies amongst retroviruses and it is tempting to speculate that this is due to differences in the affinity of such chromatin binding sequences for chromatin . Further work is needed to assess whether p12 function is a conserved general feature of retroviral replication .
For bacterial expression , N-MLV p12 sequences ( WT , Mut6 and Mut14 ) were amplified from pCIG3N [7] and cloned into pGEX6 . 1 using BamHI/XhoI sites . C-terminally His-tagged N-MLV CA WT and P1G mutant proteins were expressed from pET22-N-MLV-CA plasmids as previously described [26] . Retroviral VLPs were synthesised by co-transfection of three plasmids: an envelope expression plasmid for vesicular stomatitis virus G protein ( pczVSV-G ) [49] , a Mo-MLV-based retroviral vector encoding LacZ ( pczLTR-LacZ ) [50] and a Gag-Pol expression plasmid for either Mo-MLV ( pKB4 ) [7] for infectivity and microscopy assays or Mo-MLV with myc-tagged p12 ( pKB4mycE ) [9] for viral Co-IP assays . The generation of p12 alanine-scanning mutations in pKB4 and pKB4mycE has been described previously [7 , 9] . The M63I , G49R/E50K , S61A/D/E changes were introduced into Mo-MLV p12 and the W390A mutation was introduced into IN by mutagenesis of pKB4 , using the QuikChange II site-directed mutagenesis kit ( Agilent technologies ) and primers shown in S3 Table . GST-p12 fusion proteins were expressed from pCAGGS/GST-derived plasmids . pCAGGS/GST carrying WT Mo-MLV p12 was a kind gift from J . Martin-Serrano [5] . Mo-MLV p12 mutants ( Mutants 6 to 15 , L-dom , p12+hCBS ) were sub-cloned into pCAGGS/GST from the Mo-MLV Gag-Pol vector , pKB4 [7] using EcoRI restriction sites . N-MLV p12 was similarly sub-cloned from pCIG3N [7] . Codon-optimised FeLV , GaLV , XMRV and KoRV p12 sequences were synthesised by GeneART ( Thermo Fisher Scientific ) and then cloned in to pCAGGS/GST using EcoRI/XhoI sites . The Mo-MLV p12 mutants M63I , G49R/E50K , D25A , S61A/D/E as well as the FeLV p12 mutants I52M and A53V were generated by site-directed mutagenesis of the pCAGGS/GST plasmids using primers shown in S3 Table . The Mo-MLV p12 CTD was amplified from pKB4 using primers: for 5’-atgaattcggagaggcaccggacc-3’ and rev 5’-tggaattcaagggggctcccgtctc-3’ , and cloned into pCAGGS/GST using EcoRI sites . To generate stable GST-p12 expressing lines , GST-p12 sequences were amplified from pCAGGS/GST plasmids and sub-cloned into a Mo-MLV-based retroviral vector encoding puromycin-resistance called pCMS28 [51] using BglII/XhoI restriction sites . Transduction vectors were synthesized by cotransfection of pczVSV-G , pKB4 and pCMS28/GST-p12 . HA-tagged IN was expressed from pCMV4/HA which was generated by cloning a codon-optimised IN sequence ( synthesised by GeneART , Thermo Fisher Scientific ) into the pCMV4-HA vector [52] using MluI/SbFI restriction sites . HEK 293T , HeLa and D17 cells ( Bishop laboratory cell stocks ) were maintained in DMEM ( Thermo Fisher Scientific ) , supplemented with 10% heat-inactivated foetal calf serum ( Biosera ) and 1% penicillin-streptomycin ( Sigma ) . The cells were stored in a humidified incubator at 37°C and 5% CO2 . VLPs were made by co-transfecting 293T cells with plasmids encoding VSV-G ( pczVSV-G ) , Mo-MLV Gag-Pol and the LacZ reporter gene ( pczLTR-LacZ ) , in an equimolar ratio . Approximately 16 h after transfection the cells were treated with 10 mM sodium butyrate for 6 h to promote transcription . VLP-containing culture supernatants were harvested 48 h post-transfection and filtered to remove cellular debris . Viral titres were quantified using a modified ELISA for reverse transcriptase activity ( Cavidi ) . Transduction vectors for stable cell line generation were produced by co-transfecting 293T cells with plasmids encoding VSV-G ( pczVSV-G ) , Mo-MLV Gag-Pol ( KB4 ) and GST-p12 , as described above . HeLa cells were transduced with these VLPs by spinoculation ( 1600 g , 2 h , 16°C ) in the presence of 4 μg/ml polybrene ( Sigma ) . From approximately 72 h after infection , cells were passaged in media containing 0 . 5 μg/ml puromycin ( Thermo Fisher Scientific ) to select for transduction . The N-terminally GST-tagged p12 fusion proteins were produced in E . coli Rosetta 2 ( DE3 ) pLysS cells ( Thermo Fisher Scientific ) from pGEX . 1-derived plasmids . The cells were grown in the presence of 1% glucose and GST-p12 expression was induced by the addition of 1 mM isopropyl-β-D-thiogalactopyranoside ( IPTG ) in mid-log phase . Cells were subsequently lysed in 50 mM Tris pH 8 , 500 mM NaCl , 0 . 5 mM TCEP , 0 . 1% Triton X-100 ( Buffer A ) in the presence of protease inhibitors ( Roche ) and incubated with Lysozyme ( Sigma Aldrich ) and Benzonase ( Sigma Aldrich ) for 1 hour at 4°C . Lysis was facilitates by sonication , x2 pulses , 5 minutes , 40% amplitude and crude lysates were centrifuged at 48 , 000 g , 45 minutes , 4°C to remove debris . The clarified lysates were applied to 1 ml GST-trap columns ( GE Healthcare ) . After washing with Buffer A , untagged-p12 was eluted from the resin by digestion with 3C precision protease ( GE Healthcare ) . The eluate was then heated at 65°C for 10 minutes and centrifuged at 40 , 000 g for 20 minutes to remove precipitates . Acetic acid ( pH ∼3 ) was added to the supernatant which was then centrifuged at 40 , 000×g for 20 minutes to remove nucleotides and DNA . The supernatant was then applied to a Superdex 75 ( 16/60 ) size exclusion column equilibrated in 200 mM Ammonium bicarbonate . Eluate fractions containing p12 were pooled and lyophilised . The purity of the protein preparations was assessed by SDS-PAGE and the concentrations were determined from the absorbance at 280 nm . C-terminally His-tagged N-MLV CA WT and P1G mutant proteins were expressed and purified as previously described [26] . Binding of purified p12 to in vitro assembled CA arrays was performed essentially as previously described [26] . Briefly , lipid nanotubes were synthesised using the tube-forming lipid , d-galactosyl-β-1 , 1′ N-nervonoyl-d-erythro-sphingosine ( GalCer ) ( Avanti ) in combination with the Ni2+-chelating lipid , DGS-NTA ( Avanti ) in a 7∶3 ratio . After mixing the lipids , residual chloroform and methanol were removed under a gentle stream of nitrogen and the lipids were resuspended with the aid of sonication in 10 mM Tris-HCl pH 8 , 10 mM KCl , 100 mM NaCl , to a final concentration of 0 . 5 mg/ml . The tubes were coated by incubating with 2 mg/ml of purified His-tagged N-MLV wild type CA or P1G CA mutant at a ratio of 1∶3 with 10 mM imidazole , for 1 hour at room temperature . To assess binding of p12 to non-hexameric CA , purified 2 mg/ml CA was also immobilised on HIS-Select Nickel Affinity beads ( Sigma ) at a 1:10 ratio . Purified p12 proteins were diluted to approximately 50 μg/ml , in dilution buffer TBS ( 10 mM Tris-HCl pH 8 , 10 mM KCl , 100 mM NaCl , 10 mM imidazole ) and 1% BSA . In each binding reaction , 200 μl of p12 was incubated with 4 μl of CA-coated lipid nanotubes or 1 μl of CA-coated beads , for 2 hours at room temperature with gentle agitation . The samples were then layered on top of a 2 ml cushion of 40% ( w/v ) sucrose in TBS and centrifuged at 34 , 000 g for 1 hour at 4°C . The supernatants were then aspirated and the pellets were resuspended in 40 μl of protein loading buffer . His-tagged CA and p12 in the pellet fractions were detected by immunoblotting . Proteins were separated by SDS polyacrylamide gel electrophoresis ( SDS-PAGE ) and transferred onto polyvinylidene fluoride ( PVDF ) membranes ( Milipore ) . The primary antibodies used in immunoblotting are included in S4 Table . Unless specifically stated , the rabbit antibody raised against a p12 peptide antigen was used to probe for p12 . Goat anti-rabbit IRDye800CW ( LI-COR , 1:5000 ) , goat anti-mouse IRDye680RD ( LI-COR , 1:5000 ) and goat anti-mouse HRP ( Pierce , 1:10 , 000 ) were used as secondary antibodies . Immuno-complexes were detected either using the Li-cor Odyssey imaging and quantitation system ( LI-COR Bioscience ) or hyperfilm processing with Fijifilm FPM-3800A developer . Titres of VLPs carrying myc-tagged p12 were normalised based on CA amount . Aliquots of VLPs were concentrated by centrifugation through a 20% ( w/v ) sucrose cushion for 2 h at 10 , 000 g , 4°C and re-suspended in protein loading buffer . Samples were then analysed by immunoblotting together with a panel of serially-diluted purified N-MLV CA of known concentration . CA amounts were interpolated from the mean band intensity measurements . For Co-IP assays , myc-tagged p12 VLPs were concentrated by centrifugation through sucrose as above and re-suspended in 1% formaldehyde in PBS for 20 min at room temperature . The cross-linking reaction was then quenched for 10 min by the addition of Tris-HCl pH 7 . 5 to a final concentration of 250 mM . The cross-linked VLPs were again spun through sucrose and re-suspended in RIPA lysis buffer ( Thermo Fisher Scientific ) supplemented with protease inhibitors ( Roche ) . Lysis was facilitated by sonication , x7 pulses , 30 s ON and 30 s OFF , in an ice bath ( Decon FS100 ) . After sonication , 4 mM MgCl2 and Pierce universal nuclease ( Thermo Fisher Scientific ) were added to the viral lysates . The lysates were then normalised based on previously-estimated CA amounts and incubated overnight with Protein G Dynabeads ( Thermo Fisher Scientific ) carrying immobilised anti-myc antibody 9E10 [10] . After approximately 16 h , the beads were washed three times , re-suspended in protein loading buffer , boiled and analysed by immunoblotting with anti-p12 and anti-CA antibodies ( S4 Table ) . HeLa or D17 cells were seeded in 24-well plates ( Corning ) at densities of 2 . 5x104 cells/well and 2x104 cells/well , respectively . Cells were infected 24 h later , with WT and mutant VLPs normalised on their RT activity and incubated at 37°C for 72 h . Cells were then lysed in Tropix Lysis buffer ( Thermo Fisher Scientific ) and frozen at -20°C . To measure LacZ activity , cell lysate was mixed with Tropix galactostar reaction mixture ( Thermo Fisher Scientific ) and luminescence was measured for 1 h at 10 min intervals on a Tecan Safire plate reader . Absolute infectious titres of VLPs for microscopy assays were determined by X-gal staining of infected HeLa cells . HeLa cells were seeded in 12-well plates ( Corning ) at a density of 5x104 cells/well and infected 24 h later with a 10-fold dilution series of VLPs by spinoculation ( 1600 g , 2 h , 16°C ) . Cells were then incubated at 37°C for 30 min prior to replacement of media with warm serum-supplemented DMEM . After 72 h , cells were washed in PBS and fixed for 10 min with 2% formaldehyde and 0 . 2% glutaraldehyde . Staining was performed overnight at 37°C in PBS with 0 . 4 mg/ml X-gal ( 5-bromo-4-chloro-3-indolyl- beta-D-galactopyranoside ( Sigma ) , 4 mM K3Fe ( CN ) 6 ( Sigma ) , 4mM K4Fe ( CN ) 6ˑ3H2O ( Sigma ) , 2mM MgCl2 ( Thermo ) . The number of blue LacZ-expressing colonies were counted using a light microscope ( Olympus ) and the viral titre calculated . To synchronise the cell cycle of HeLa cells , 2 μg/ml of aphidicolin ( Sigma ) was added for 15–16 h . The cells were then washed and incubated in normal media for 0 . 5–1 . 5 h before re-seeding on 13 mm coverslips . Cells were incubated at 37°C for 7–8 h before the addition of 2 μg/ml of aphidicolin for a further 14–15 h . Cells were then washed and allowed to recover for 30 min in normal media before infection with VLPs by spinoculation ( 1600 g , 2 h , 16°C ) at a MOI of <1 . The cells were incubated at 37°C for 30 min before replacement of media with fresh serum-supplemented DMEM and returned to the incubator for 10 h . The cells were then washed twice with PBS and fixed with 4% paraformaldehyde for 5 min at room temperature followed by ice-cold methanol at -20 oC for 5 min . Cells were subsequently permeablised with 0 . 5% saponin ( Sigma ) in PBS for 30 min and blocked in 5% normal donkey serum ( NDS , Source Bioscience ) and 0 . 5% saponin in PBS for at least 1 h . Cells were then incubated with rabbit anti-p12 ( custom generated against a p12 peptide by Cambridge Research Biochemicals ) and rat anti-CA ( CRL-1912 , ATCC ) antibodies diluted in 1% NDS and 0 . 5% saponin in PBS for 1 h , at RT . Coverslips were then washed three times with PBS and incubated for another hour with anti-mouse and anti-rat secondary antibodies conjugated to Alexa Fluor 594 and 488 dyes ( Abcam , ab150064 and ab150117 ) , diluted in 1% NDS PBS and 0 . 5% saponin in PBS . Coverslips were finally washed three times and mounted in Prolong Gold media with DAPI ( Thermo Fisher Scientific ) on glass slides ( Menzel-Gläser ) . Clear nail varnish was used for sealing the coverslips . Immunostained cells were visualised on a SP5 inverted confocal microscope ( Leica ) using a HCX PL APO CS 100 . 0x1 . 46 OIL ( Leica ) objective . Image analysis was performed using FIJI https://fiji . sc . For co-localisation analysis , three-dimensional stacks were projected in Z by summation to generate two-dimensional images . The DAPI channel was then pre-processed by median filtering to suppress noise and grey-level thresholded using the Huang algorithm to generate binary mask images . Interphase nuclei and artefacts were removed by specifying object area and circularity thresholds in the Particle Analyzer . Co-localisation analysis of p12 and CA puncta was then performed using a custom plug-in produced in-house–further details and complete source code are available online ( https://bitbucket . org/djpbarry/particletracker/wiki/Particle%20Mapper ) . Briefly , puncta in one channel were detected using a Laplacian-of-Gaussian based detection scheme ( referred to as ‘Blobs’ detection mode within the plug-in ) –puncta exhibiting an intensity below a pre-specified threshold were discarded . Co-localisation was then evaluated by determining how many puncta in the first channel are co-incident with local intensity maxima above a second threshold exist in the second channel . HeLa cells stably expressing GST-p12 were fixed in 4% paraformaldehyde and methanol as described above and then stained with rabbit anti-p12 or anti-GST ( Abcam , ab19256 ) primary antibodies and the Alexa Fluor 594 secondary antibody . The samples were visualised on SP5 using a 100X ( 1 . 46NA ) objective and analysed on ImageJ 1 . 49v . GST-p12 proteins were expressed in 293T cells from pCAGGS/GST-derived plasmids by transient transfection using Turbofect ( Thermo Fisher Scientific ) . Cell media was changed 24 h after transfection , and cells were incubated overnight at 37°C , before being harvested , counted , pelleted ( 500 g , 5 min ) and snap frozen in dry ice/ethanol . Biochemical fractionation of cells was performed essentially as described in [14] . Briefly , cell pellets thawed on ice were resuspended at 2 . 5× 107 cells/ml in buffer 1 ( 10 mm HEPES pH 7 . 9 , 10 mm KCl , 1 . 5 mm MgCl2 , 10% glycerol , 0 . 34 m sucrose , 1 mm DTT and protease inhibitors ) and then supplemented with 0 . 1% Triton-X-100 . Cells were incubated on ice for 5 min and nuclei ( P1 ) were collected by centrifugation at 1300 g , 5 min , 4°C . The supernatant ( S1 ) was clarified by further centrifugation ( 5 min , 20 000 × g , 4°C ) to collect the cytosolic supernatant fraction ( S2 ) . The P1 nuclei were washed once with buffer 1 and then lysed for 30 min in buffer 2 ( 3 mM EDTA , 0 . 2 mM EGTA and 1 mM DTT , protease inhibitors ) . The chromatin ( P3 ) and soluble nuclear ( S3 ) fractions were separated by centrifugation ( 5 min , 1700 g , 4°C ) . The S2 , S3 and P3 fractions were analysed by immunoblotting . GST-p12 proteins were expressed in 293T cells from pCAGGS/GST-derived plasmids by transient transfection using Turbofect ( Thermo Fisher Scientific ) . Approximately 24 h after transfection , cells were changed into fresh media containing either 400 ng/ml nocodazole ( Sigma ) or 2 μg/ml aphidicolin ( Sigma ) and incubated overnight at 37°C . The synchronisation of the cell cultures after drug treatment was assessed by propidium iodide staining and cell cycle phase analysis by flow cytometry . Cells to be used in pull down assays were washed and lysed in 20 mM Tris pH 8 . 0 , 300 mM NaCl , 1 mM EDTA , 10% Glycerol , 1% Triton X-100 , protease inhibitors ( Roche ) , phosphatase inhibitors ( Merck ) . Lysis was facilitated by passing the lysates through a 19-gauge needle x7 and sonication , x7 pulses , 30 s ON and 30 s OFF , in an ice bath ( Decon FS100 ) . After sonication , the lysates were supplemented with 4 mM MgCl2 and Pierce universal nuclease ( Thermo Fisher Scientific ) and incubated for a further 1 h at 4°C . The lysates were then clarified by centrifugation and normalised based on total protein concentration using the Pierce BCA Protein Assay kit ( Thermo Fisher Scientific ) . 0 . 5 ml aliquots of lysates at 1 . 5–3 mg/ml were incubated with glutathione-sepharose beads ( 100 μl/reaction of a 50% slurry ) ( GE Healthcare ) for 3 h at 4°C with end-over-end rotation . The beads were then washed three times in 20 mM Tris pH 8 . 0 , 300 mM NaCl , 1 mM EDTA , 10% Glycerol , 0 . 1% Triton X-100 , protease inhibitors and re-suspended in 50 μl of x2 protein loading dye for SDS-PAGE and immunoblotting . Silver-staining was performed using the SilverQuest kit ( Thermo Fisher Scientific ) . ProQ diamond and Sypro ruby staining ( Thermo Fisher Scientific ) of gels was visualised and quantified on a ChemiDoc imaging system ( Bio-Rad ) . For SILAC-MS experiments , 293T cells were grown is media containing light ( R0K0 , 12C-Arginine and 12C-Lysine ) , medium ( R6K4 , 13C-Arginine and Lysine-4 , 4 , 5 , 5-d4 ) or heavy ( R10K8 , 13C+15N-Arginine and 13C+15N-Lysine ) versions of L-Arginine and L-Lysine amino acids ( Pierce ) for a minimum of 6 doublings before transient expression of GST-p12 proteins . After performing the glutathione-sepharose pull-down assays equivalent volumes of the beads eluates were pooled and submitted to the Bristol University Proteomics Facility for RP nano-LC-MS/MS analysis . This analysis was performed on an LTQ-Orbitrap Velos mass spectrometer in line with a Dionex Ultimate 3000 nanoHPLC system as described in [53] . To identify phosphopeptides of GST-p12 , bead eluates were separated by SDS-PAGE and stained using SimplyBlue safe stain ( Thermo Fisher Scientific ) . The band corresponding to GST-p12 was then excised and submitted for nano-LC-MS/MS analysis as described in [54] , but without phosphopeptides enrichment . The probability-based phosphoRS algorithm was used for assigning phosphorylation sites in the detected peptides [32] . GST-p12 and IN-HA were expressed in 293T cells from pCAGGS/GST and pCMV4/HA-derived plasmids respectively , by transient transfection using Turbofect ( Thermo Fisher Scientific ) . Approximately 24 h after transfection , cells were changed into fresh media and incubated overnight at 37°C . Cells were then washed and lysed in 20 mM Tris-HCl pH 8 . 0 , 300 mM NaCl , 1 mM EDTA , 10% Glycerol , 1% Triton X-100 , protease inhibitors . Lysis was facilitated by passing the lysates through a 19-gauge needle x7 and sonication , x7 pulses , 30 s ON and 30 s OFF , in an ice bath ( Decon FS100 ) . After sonication , the lysates were supplemented with 4 mM MgCl2 and 1 mM DTT , and incubated for a further 1 h at 4°C . The lysates were then clarified by centrifugation and diluted 1:3 in 20 mM Tris-HCl pH 8 . 0 , 10% glycerol , 4 mM MgCl2 and 1 mM DTT . The lysates were then normalised based on total protein concentration using the Bradford protein assay ( Bio-Rad ) . 0 . 7 ml aliquots of lysates at 0 . 7 mg/ml were incubated with calf thymus DNA-coated cellulose beads ( 100 μl/reaction of a 50% slurry ) ( Sigma ) for 1 h at 4°C with end-over-end rotation . The beads were then washed three times in 20 mM Tris pH 8 . 0 , 100 mM NaCl , 0 . 33 mM EDTA , 10% Glycerol , 0 . 33% Triton X-100 , 4 mM MgCl2 and 1 mM DTT and re-suspended in 50 μl of x2 protein loading dye for SDS-PAGE and immunoblotting . GST-p12 proteins were expressed in 293T cells from pCAGGS/GST-derived plasmids by transient transfection using Turbofect ( Thermo Fisher Scientific ) . Approximately 24 h after transfection , cells were changed into fresh media containing 400 ng/ml nocodazole ( Sigma ) and incubated overnight at 37°C . Cells were then changed into media containing 400 ng/ml nocodazole , 10 μM MG132 ( Sigma ) and either 40 mM LiCl ( Sigma ) , 75 μM roscovitine ( Sigma ) or 40 μM kenpaullone ( Sigma ) , for 3 . 5 h . Cells were immediately washed and lysed for glutathione-sepharose pull-down assays as described above . The binding of PFV hCBS and p12 CTD peptides to recombinant H3 . 3 poly-nucleosomes ( Active Motif ) was measured using biolayer interferometry on an Octet RED system ( Pall ForteBio Corp ) . The peptides were synthesised with a N-terminal biotin tag and GGGG linker . The Mo-MLV , FeLV and GaLV p12 CTD peptide sequences used are highlighted in bold in Fig 5C . A negative control was included that contained an R66A change in the Mo-MLV p12_M63I CTD peptide . Mo-MLV phos p12_M63I CTD was also synthesised with a phosphorylated S61 residue . The hCBS ( positive control ) sequence used corresponds to the CBS of PFV Gag ( NQGGYNLRPRTYQPQRYG ) . The biotinylated peptides were loaded onto streptavidin biosensors ( Pall ForteBio Corp ) at 0 . 5 μg/ml for 120 s in 10 mM HEPES pH 7 . 5 , 150 mM NaCl , 0 . 005% ( v/v ) Tween 20 . Following a buffer wash , the biosensors were incubated with serially-diluted nucleosome preparations ( ~0 . 5–250 nM ) for 900–5400 s to measure association . Experiments were performed at 25°C and sample plates were agitated at 1000 rpm . Equilibrium binding amplitudes were determined by exponential least squares curve fitting of the association phase . The estimated binding amplitudes were then plotted against approximate nucleosome concentrations on GraphPad Prism 7 . Equilibrium binding constant ( Kd ) and maximum specific binding ( Bmax ) were calculated by fitting the data to a non-linear regression model . Experiments were performed at least in duplicate and the data pooled for fitting . For competition assays , soluble non-biotin tagged peptides were added to the nucleosomal arrays at 100 μM to inhibit nucleosome binding prior to recording BLI measurements with biotinylated peptide immobilised on sensor probes . The non-biotin tagged peptide sequences were: hCBS—NQGGYNLRPRTYQP; hCBS_L539A/R540A - NQGGYNAAPRTYQP; p12_WT—DPSPMASRLRGRREPPY; p12_R66A - DPSPMASALRGRREPPY .
|
In addition to matrix , capsid and nucleocapsid , the Gag polyproteins of many retroviruses also encode additional cleavage products , such as the p12 protein of murine leukemia virus . p12 is essential for both early and late replication events , but its function during early infection remains poorly characterised . Recent work has shown that the N-terminal domain of p12 binds to and stabilises the capsid shell that surrounds the viral core . Defects in the C-terminal domain of p12 prevented the viral pre-integration complex ( PIC ) from localising to chromatin during infection , and this could be rescued by the addition of a heterologous chromatin binding domain . In this study , we show that p12 is able to tether viral PICs containing capsid to mitotic chromatin by directly binding both hexameric capsid and nucleosomal histone proteins . Additionally , the affinity of p12 for chromatin may be enhanced by its binding to capsid and by its phosphorylation in mitosis . Excitingly , the mechanism of nucleosome binding appears to be conserved between gammaretroviruses and spumaretroviruses . Furthermore , influencing global nuclear targeting of the PIC by linking capsid to chromatin is reminiscent of the proposed role of the capsid-binding host protein , CPSF6 , in HIV-1 infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"phosphorylation",
"molecular",
"probe",
"techniques",
"293t",
"cells",
"cell",
"cycle",
"and",
"cell",
"division",
"biological",
"cultures",
"cell",
"processes",
"immunoblotting",
"dna-binding",
"proteins",
"mitosis",
"epigenetics",
"molecular",
"biology",
"techniques",
"chromatin",
"research",
"and",
"analysis",
"methods",
"chromosome",
"biology",
"proteins",
"gene",
"expression",
"recombinant",
"proteins",
"histones",
"cell",
"lines",
"molecular",
"biology",
"nucleosomes",
"biochemistry",
"cell",
"biology",
"post-translational",
"modification",
"genetics",
"biology",
"and",
"life",
"sciences"
] |
2018
|
Murine leukemia virus p12 tethers the capsid-containing pre-integration complex to chromatin by binding directly to host nucleosomes in mitosis
|
Chronic wounds are a significant socioeconomic problem for governments worldwide . Approximately 15% of people who suffer from diabetes will experience a lower-limb ulcer at some stage of their lives , and 24% of these wounds will ultimately result in amputation of the lower limb . Hyperbaric Oxygen Therapy ( HBOT ) has been shown to aid the healing of chronic wounds; however , the causal reasons for the improved healing remain unclear and hence current HBOT protocols remain empirical . Here we develop a three-species mathematical model of wound healing that is used to simulate the application of hyperbaric oxygen therapy in the treatment of wounds . Based on our modelling , we predict that intermittent HBOT will assist chronic wound healing while normobaric oxygen is ineffective in treating such wounds . Furthermore , treatment should continue until healing is complete , and HBOT will not stimulate healing under all circumstances , leading us to conclude that finding the right protocol for an individual patient is crucial if HBOT is to be effective . We provide constraints that depend on the model parameters for the range of HBOT protocols that will stimulate healing . More specifically , we predict that patients with a poor arterial supply of oxygen , high consumption of oxygen by the wound tissue , chronically hypoxic wounds , and/or a dysfunctional endothelial cell response to oxygen are at risk of nonresponsiveness to HBOT . The work of this paper can , in some way , highlight which patients are most likely to respond well to HBOT ( for example , those with a good arterial supply ) , and thus has the potential to assist in improving both the success rate and hence the cost-effectiveness of this therapy .
Chronic leg ulceration is a significant socioeconomic problem [1] . Those who suffer from leg ulcers experience considerable pain , immobility and decreased quality of life [2] . Approximately 3% of the over 60 age group suffer from lower limb ulceration [3] . A successfully healing wound ( or an “acute” wound ) is typically thought to progress through four stages; haemostasis , inflammation , proliferation and remodelling [4] , [5] , although these processes are interconnected and overlapping . Haemostasis should last a matter of hours during which time the blood flow is stopped . Inflammation sees the production of chemoattractants that stimulate fibroblasts , the dominant cell in the proliferative stage of healing , to migrate into the wound site and to produce collagen , the main component of the extracellular matrix ( ECM ) . The cocktail of chemoattractants also stimulate the systematic rearrangement of endothelial cells ( ECs ) from neighbouring blood vessels [6] . Capillary sprout extension is facilitated by EC proliferation and further migration toward the chemical attractant . The joining of two capillary sprouts within a healing wound forms a loop through which blood can flow and new sprouts develop from this vessel thus propagating angiogenesis [7] . A chronic wound is one in which healing fails to proceed through an orderly and timely process to produce anatomic and functional integrity , or proceeds through the repair process without establishing a sustained anatomic and functional result [8] . The factors responsible for the development of a chronic wound remain unclear , however the most common cause , according to Mathieu [9] , is thought to be related to the detrimental effects of prolonged wound hypoxia ( oxygen deficiency ) . Enoch et al . reports that chronic wounds can be arrested in any one of the stages of wound healing , but disruption commonly occurs in the inflammatory or proliferative phases [4] . HBOT involves the intermittent exposure of the body to 100% oxygen at a pressure greater than 1 atmosphere ( atm ) and its use is supported in the treatment of problem wounds [10] . However , there is much debate about the optimal HBOT protocol in treating such wounds [11]–[14] . Although HBOT is typically used as an adjunctive therapy for treating chronic wounds , many clinicians lack a full knowledge of the evidence-based data that support its use [15] . The primary rationale behind the use of HBOT in the treatment of chronic wounds is to elevate the amount of oxygen delivered to the wound site [16] . For a more detailed review of the wound healing process , the different etiologies of chronic wounds and the use of HBOT to treat nonhealing wounds see Thackham et al . [5] . It should be noted that HBOT is not the only wound healing therapy currently being studied . Gordillo et al . review the use of topical oxygen therapy to assist the closure of chronic wounds [17] . While the wound healing process is undeniably complex , there are useful mathematical models that address various aspects of the phenomenon . The models can be categorised into four broad groups: continuum reaction-diffusion models , mechanochemical models , discrete/stochastic models and multiscale models . Discrete models have the ability to contain a level of detail that is not possible from a continuum model and , in general , allow for quicker numerical simulation however , continuum models allow mathematical analysis that discrete models do not . Continuum reaction-diffusion models are arguably the most commonly used theoretical approach for studying the angiogenesis process . One of the first models of this kind is due to Balding and McElwain , who developed a model to investigate tumour-induced angiogenesis [18] . Their model consisted of distinct equations for the blood vessel and capillary tip density . Edelstein had previously used a similar approach to model fungal growth [19] . The concept behind including both blood vessel and tip species is that the ECs in the tip of a vessel guide the ECs in the sprout . This aspect of the model , the so-called “snail-trail” production of blood vessels , means that if the capillary tip density , , moves with a velocity , then the rate of increase ( that is , production/extension ) of blood vessels is given by , where is a unit vector in the direction of . The model by Balding and McElwain accounts for branching and anastomosis but does not account for the extension of vascular loops . In 1996 , Pettet and coworkers developed two models of wound healing angiogenesis [20] , [21] . Sherratt , in 2002 , described the work in [21] as “the most important theoretical work on wound angiogenesis to date” . In [20] Pettet et al . proposed a three species simplification of the more detailed six species model of wound healing angiogenesis presented in [21] and used an analytic approach to obtain an approximate solution . The first and foremost feature of these models is that they incorporate the dependence of chemoattractant production on the local wound oxygen concentration , with chemoattractant production occurring in a specific oxygen concentration range . Both of these publications modelled the extension of blood vessels using the Balding and McElwain “snail-trail” concept . Other authors have chosen this approach to model wound and tumour angiogenesis ( see for example [7] , [22] , [23] ) . In 2002 , Gaffney et al . used a two-species model to investigate cutaneous wound healing [24] . Here a travelling wave analysis was used to identify a lower bound on the wave speed of the wound healing unit in terms of two key model parameters , namely , the random motility of capillary tips and the rate of budding of tips . Importantly , Gaffney et al . chose not to use a snail-trail approach , opting rather to consider the EC density explicitly in addition to the capillary tip density [24] . The flux of the capillary tips is determined by random motion and directed motion . The flux of the EC is then assumed to be proportional to the tip flux , where the rate constant is the number of EC that makes up an average capillary tip . More recently , Schugart et al . developed a seven species model of acute wound healing angiogenesis [25] , using a similar approach to Gaffney et al . [24] in their treatment of the EC flux , while Addison-Smith et al . used a simple mechanistic model for the sprouting of vessels during tumour-induced angiogenesis [26] . Mantzaris et al . provides an excellent review of continuum models of angiogenesis and concludes that continuum models are important for providing significant insight into the relative importance of different processes [27] . The interactions between cells and the substratum in wound healing are not just chemical; there is also a mechanical influence [27] . For instance , in order to migrate , ECs extend lamellipodia in the direction of migration , and exert tractional forces on the ECM . Mechanochemical models of wound healing are essentially continuum models that account for the forces that cells exert on the ECM . Mechanochemical models of wound healing include the early work by Tranquillo and Murray [28]–[30] and the extensions by Olsen et al . and Cook [31] , [32] . These models consider the connection of cells to the ECM and are thus relevant for deeper , dermal wounds . They are typically applied to acute ( normal ) healing wounds that heal primarily by contraction . Mechanochemical modelling is needed when considering the remodelling phase of wound healing since it is during this stage that the wound contracts . However , since we are interested in chronic wound healing , and chronic wounds typically arise due to complications in the inflammatory or proliferative phase , the use of mechanochemical models is not addressed here . Furthermore , healing in human wounds is predominantly due to proliferation and migration of cells from outside the wound , whereas in animal models , the contraction of the wound by mechanical forces is thought to be more substantial . It is interesting to note that while the critical role of oxygen in wound healing is well known there are no mechanochemical models , to date , that incorporate angiogenesis . Discrete models , using cellular automata for example , have been used to capture key features of angiogenesis including the outgrowth , branching and anastomosis of vessels . Stokes and Lauffenburger pioneered the discrete modelling of blood vessel formation in their series of publications [33]–[35] . Their work is discrete in that they present a stochastic model for the random motility and chemotaxis of individual cells . Other aspects of their model use continuum modelling , for example , in the conservation equation for the chemoattractant . Levine and coworkers extended this approach with a series of publications [36]–[40] in which they investigated the mathematical modelling of tumour-induced angiogenesis . In these models , continuous limits of reinforced random walk equations govern the angiogenesis process while ordinary differential equations ( ODEs ) model the biochemical kinetic equations . In their review , Mantzaris et al . states that one main advantage of using discrete models rather than continuum ones is that individual cells and sprouting of vessels can be tracked [27] . Although discrete models can be computationally fast and efficient , and provide quantitative numerical data , these models are not as readily amenable to mathematical analysis as continuum models are . Multiscale techniques have been used to simulate the wound healing process ( see for example Dallon et al . and Cai et al . [41] , [42] ) . Sun et al . have developed several models of angiogenesis including a multiscale model where the concentration of chemoattractant is modelled at the tissue scale , while the capillary network is modelled at the cellular scale [43] , [44] . More recently , Alarcon and coworkers and McDougall and coworkers have used multiscale techniques to investigate angiogenesis associated with tumour growth [45]–[48] . The overall aim of this paper is to use a theoretical model to evaluate the use of hyperbaric oxygen therapy as an adjunct therapy ( a therapy used to assist a primary treatment ) for treating chronic wounds . Through numerical simulations we conclude that intermittent hyperbaric oxygen therapy has the potential to aid in the healing of chronic wounds ( a chronic wound is one which does not heal in an orderly set of stages and in a reasonable amount of time in the way that most wounds do ) .
We simulate an acute healing wound with the choice of parameters outlined in Table 1 , noting that this choice of values yields a steady state oxygen concentration behind the wave front ( behind the healing front the oxygen concentration , , tends to as ) above the lower threshold for capillary tip production , , so that healing will be initiated . Fig 1 shows such a normal situation in which a wound of length 2 cm ( that is , ) is almost completely reoxygenated within 2 weeks . It would take longer than this , roughly 2 . 5 weeks , for the simulated wound to completely revascularise . We note that the vessel density can rise above the carrying capacity , , due to rapid chemotaxis and may remain elevated until the remodelling process drives the density to return to normal levels . A chronic wound is simulated by selecting parameter values such that the oxygen concentration behind the injured tissue ( near the oxygen concentration tends to ) does not rise above the lower threshold for capillary tip production , . As mentioned above , our assumption is that chronicity is associated with a reduced or impeded supply of oxygen from the vasculature . We therefore reduce the value of used for the simulation in Fig 1 by a factor of 10 to . The resulting simulation produces an oxygen profile that is always within the range ( that is , below the lower threshold for capillary tip production ) inside the wound . Fig 2 shows the chronic wound simulation . We note no significant change over time , indicating that no healing is taking place . 10 . 1371/journal . pcbi . 1000451 . g002Figure 2 Simulation of a chronic wound in which no healing occurs . Multiple day intervals are shown ( dark blue = 2 , red = 4 , green = 6 , black = 8 , yellow = 10 , light blue = 12 , pink = 14 ) . Parameter values: as per Fig 1 , except . We now investigate the impact of HBOT on the healing of a chronic wound . The strength and duration of HBOT are given by the parameters and in the model , respectively . The parameter is a measure of the relative increase in supply of oxygen during HBOT compared to times of no treatment . Fig 3 shows such a chronic wound situation under HBOT with for hours per day ( that is , of a day ) . A value of is associated with 100% oxygen at a pressure of just under 3atm ( see Materials and Methods Section ) , which is a reasonable treatment protocol . We note from the simulation that the capillary tip density in the chronic wound reaches highly elevated levels under treament and that healing is quickly initiated in the chronic wound . Table 2 shows that an value of 5 . 73 equates to 100% oxygen at 1atm ( that is , normobaric oxygen therapy ) . The analysis presented later ( see Expression ( 22 ) in “Analysis of Feasible HBOT Protocol” ) predicts that a chronic wound simulated with will not heal and this is confirmed by numerical simulations . Thus we have shown , under the assumptions on the model presented here , that normobaric oxygen will not stimulate healing of a chronic wound and we have answered the somewhat controversial question of whether or not normobaric oxygen can be used to substitute for HBOT in the treatment of chronic wounds . Normobaric oxygen fails to stimulate healing in the chronic wound since the oxygen levels under the treatment are still insufficent to initiate capillary tip production . Similarly , our numerical simulations reveal that values of in excess of about 2000 are too high to enable healing to occur . This is because the oxygen levels are raised so much under the treatment that capillary tip production is switched off when the upper oxygen threshold , , is reached and surpassed . Such high values of are not considered physically feasible ( see Table 2 ) . The results presented in Fig 4 reveal what happens when we simulate a situation in which HBOT is halted prematurely ( after 5 days ) . Interestingly , the effects of HBOT seemed to persist for some time after treatment is halted , but the healing progress slows considerably ( compare Figs 3 and 4 ) . Thus , if we want the wound to close as quickly as possible , then HBOT should not be terminated until complete healing of the wound is observed . Note that this is in disagreement with typical clinical protocols , which is to apply the therapy daily for about 6 weeks [16] . This restriction is likely based on cost considerations rather than clinical or experimental evidence which indicates that this is more effective in stimulating healing than continuing until the wound is completely healed . Many hyperbaric centers around the world advocate the use of HBOT to treat ‘normal’ wounds on the basis that HBOT may accelerate healing in sports injuries [49] . This use of HBOT is highly controversial [50] . Typically a sports injury is internal ( muscular ) rather than dermal , but here we consider the effect of applying HBOT to a normal healing wound . Comparing Figs 1 and 5 , we see that there is little benefit ( at most , a 10% increase in the rate at which blood vessels are progressing through the wound space ) in applying HBOT to a wound that progresses through the healing process of its own accord . Furthermore , the relatively high-cost of HBOT further detracts from the appeal of its use to treat such wounds . This simulation also shows that the capillary tip density falls significantly towards the end of the healing process . Numerical experimentation reveals that healing will occur , even with very small levels of capillary tips , suggesting that it is the presence of capillary tips , rather than their quantity , that is important for initiating healing . Clinically , this means that stimulating capillary tip production is the crucial factor that enables a chronic wound to heal when HBOT is applied . We note from Fig 1 and 5 that the density of capillary tips at comparable times is lower in the treated wound than the untreated one . This result can be explained as follows . The normal wound develops a vasculature which supplies sufficient oxygen to the wound to initiate angiogenesis . The wound does not require HBOT to heal . The application of HBOT to this wound increases oxygen levels and reduces the net tip production during times of treatment , resulting in a decreased capillary tip density . In the preceding paragraph , we deemed the therapy to have a positive effect on healing however , this was based on the faster progression of blood vessels through the wound site under the treatment . We now discuss the potential clinical implications of the restrictions derived for the HBOT protocol , the mathematical detail of which is shown in “Analysis of Feasible HBOT Protocol” in “Materials and Methods” . By considering the feasible HBOT protocol ( that is , the range of values , where is the relative increase in supply of oxygen during HBOT ) to be those that where and are the lower and upper oxygen concentration thresholds , respectively , for capillary tip production to take place , we are able to derive constraints that depend on key parameter values from the model for the range of feasible HBOT regimes . We note that if a lower bound is too high then a patient would need to be exposed to levels of pressure that are not safe in order for healing to be observed . For instance , if a particular set of wound parameter values lead to a lower bound that exceeds twenty-one , then we must conclude that HBOT will not assist this patient , since only is clinically reasonable ( see Table 2 ) . We consider two approaches to estimating the feasible protocol region . The first approach is to assume that the kinetics dominate the evolution of the oxygen concentration within the wound space and the second is to assume that the blood vessels do not migrate into the wound significantly over the first 24 hours of healing and to solve the resulting partial differential equation ( PDE ) for the oxygen concentration using Green's functions . We will not consider the upper bound from either approach since both failed to yield clinically relevant restrictions . Instead we focus on the two lower bounds , namely: ( 1a ) and ( 1b ) where and and are the lower bounds for the first and second approach , respectively , described above . We use Eqs ( 1a ) and ( 1b ) to identify patients who are unlikely to benefit from HBOT: these individuals will have higher values of and . Examining Eqs ( 1a ) and ( 1b ) reveals that patients with the following characteristics are unlikely to benefit from HBOT:
In this paper we have developed a simple mathematical model that simulates the healing of both acute and chronic wounds . The modelling framework is based on the premise that chronic wound healing is associated with poor and/or impeded oxygen delivery from the vasculature . We are now in a position to extend the model to investigate other hypotheses . For example , chronic wound healing may be associated with impaired cellular function such as poor cell chemotactic responsiveness [55] . Alternatively , certain chronic wounds may be extremely hypoxic because there is a high bacterial load in the wound bed [56] . Our model can easily be adapted to study this situation by increasing , the rate of removal/consumption of oxygen in the equation governing the oxygen distribution , Eq ( 2a ) . The development of new blood vessels occurs by two processes , namely , angiogenesis and vasculogenesis . Here we have only modeled the effect of oxygen on angiogenesis . The extension of the model to include the vasulogenesis and its potential as a mechanism for the improved healing associated with HBOT is a further extension of the model . There has already been some work on the role of vasculogenesis in tumour growth ( see for example Stamper et al . [57] ) . We used our model to evaluate the effect of treating chronic wounds with HBOT . In summary , our simulations have allowed us to make several , clinically-relevant conclusions including the following: By considering the feasible HBOT protocol ( that is , the range of values , where is the relative increase in supply of oxygen during HBOT ) to be those that where and are the lower and upper oxygen concentration thresholds , respectively , for capillary tip production to take place , we were able to derive constraints that depend on key parameter values from the model for the range of feasible HBOT regimes . By considering patients that will need excessive exposure to pressure in order to stimulate healing in conjunction with the lower bounds on the parameter , we predict that patients with any of the following conditions are unlikely to respond well to HBOT: In conclusion , we have used a simple three species model of wound healing to evaluate the effect of treating chronic wounds with HBOT . While the causal reasons for the improved healing remain unclear , protocols will remain empirical and an unreliable screening process for appropriate patients will remain in place . The work of this paper is a first step towards identifying in a systematic manner patients who are likely to respond well to HBOT and thus has the potential to assist in improving both the success rate and the cost-effectiveness of this therapy .
We now determine parameter constraints on the HBOT protocol by considering the change in oxygen concentration over the first 24 hours of treatment . Consideration of Eqs ( 2a ) – ( 2c ) reveals that unless the oxygen concentration somewhere within the wound space falls in the range , then production of capillary tips is not possible and healing will not occur . We thus define “feasible” values to be those that Current clinical protocol is to administer 1 . 5 hours of treatment once per day . Hence the above two conditions can be expressed mathematically as: We consider two approaches to deriving the aforementioned constraints . The first involves assuming that the kinetics of Eq ( 2a ) dominate the oxygen concentration within the wound space . This allows us to consider a time-dependent ODE for the oxygen concentration which is solved using standard techniques . The second approach is to assume that the blood vessels do not migrate into the wound significantly over the first 24 hours of healing . The resulting partial differential equation ( PDE ) for the oxygen concentration decouples from the remaining equations and can be solved using Green's functions [90] . By assuming that the kinetics dominate the evolution of oxygen within the wound space we arrive at the following ODE that governs the transition from the steady state oxygen concentration without HBOT to the steady state concentration under treatment ( 6 ) where , represents the proportion of each day that a patient is administered HBOT ( that is , 90 minutes per day ) . From Eq ( 2a ) , the blood vessel density at steady state is . Substituting this into Eq ( 6 ) we obtain ( 7 ) with . Eq ( 7 ) has solution ( 8 ) Using Eq ( 8 ) it is straight forward to show that our conditions for the feasible HBOT protocol as outlined previously can be written: These inequalities identify a region of values ( that is , HBOT protocols ) for treating a chronic wound such that pro-healing effects are predicted ( under the given assumptions ) to be observed: ( 9 ) where capillary tip production can only take place in the oxygen range of , is the rate at which oxygen is removed from the wound via the vasculature , is the characteristic blood vessel density , is the rate at which oxygen is supplied by the blood vessels and is the initial oxygen concentration at the wound edge . It should be noted that this analysis is based on the assumption that the wound is chronic in that , if it is left untreated , then the steady state oxygen concentration will not rise above the lower threshold for capillary tip production ( that is , ) . By substituting the estimated parameter values shown in Table 1 into the inequality in ( 9 ) , we predict that for this particular set of parameter values , HBOT will assist healing if:Recall that the parameter is dimensionless . It represents the increase in supply of oxygen during HBOT relative to periods without the treatment . It should also be noted that the upper limit of this region of feasible values is outside what would be considered “clinically reasonable” . Numerical experimentation reveals that a chronic wound exposed to HBOT with will not heal , which is in violation of the inequalities predicted by the above analysis . There are a number of potential reasons why our lower bound does not provide an accurate restriction on the HBOT protocol including: To emphasize the fact that the steady state approach is inappropriate for deriving the constraints on the feasible HBOT regime , let us consider a chronic wound exposed to HBOT with . Fig 6 compares the oxygen concentration at the wound margin ( that is , at ) during the first day of healing with the value predicted from the above steady state analysis . Note from Fig 6 that the concentration of oxygen at the wound edge differs substantially from the value associated with the steady state analysis . On closer inspection Fig 3 reveals that during the first day , the blood vessels do not migrate deep into the wound . By assuming that the blood vessel distribution throughout the wound domain does not change from the initial conditions during the first day of treatment then the oxygen PDE decouples and we have: ( 10 ) where and subject to the boundary conditions . We find the solution to Eq ( 10 ) over the first day of treatment using Green's functions:where is the initial distribution of oxygen ( see Eq ( 3d ) ) , is given above , andEvaluating the integrals gives: ( 11 ) where , is the duration of the HBOT session on the first day of treatment and is the Heaviside function . By imposing we find: ( 12 ) and similarly , we can ensure that if does not exceed: ( 13 ) where and . In terms of the estimated parameter values , shown in Table 1 , the region of feasible predicted by using Green's function is given by: ( 14 ) This lower bound is consistent with numerical simulations , which reveal that healing does not occur with . However , we note that the upper limit is significantly higher than what would be considered clinically reasonable . In Fig 7 we compare the oxygen concentration at the wound margin during the first day of healing with that predicted using the Green's function analysis . We note that there is excellent agreement between the values obtained by solving the full model numerically and by using the Green's function approach . It should be noted that in practice only the lower bounds presented here are useful since exposing a patient to high levels of oxygen for even short periods of time causes oxygen toxicity [91] . In fact , 100% oxygen for 3 hours at 3 atm can cause central nervous system breakdown [50] . Hence a value of greater than 20 . 93 ( see Table 2 ) should be considered clinically irrelevant .
|
In the time it takes you to read this paragraph , one person will have undergone a lower limb amputation due to diabetic foot disease . With the global diabetes population on the rise and set to reach 330 million by 2025 , the need for research into therapies and technologies that have the potential to prevent amputation is dire . There is much debate about the best way to treat these wounds , and one treatment that is shrouded with controversy is Hyperbaric Oxygen Therapy ( HBOT ) . There are currently no conclusive data showing that HBOT can assist chronic wound healing , but there has been some clinical success . In light of how expensive properly designed clinical trials can be , we must turn to alternative methods of assessment , such as the theoretical model presented here . The mathematical model reproduces a number of clinical observations . A key result is that while HBOT can assist chronic diabetic wounds , it holds little benefit for wounds that would heal of their own accord . This model represents a useful tool to analyse the optimal protocol , and the results and insights gained from the model may be used to improve both the success rate and thus the cost-effectiveness of this therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"mathematics",
"diabetes",
"and",
"endocrinology",
"computational",
"biology/systems",
"biology"
] |
2009
|
A Three Species Model to Simulate Application of Hyperbaric Oxygen Therapy to Chronic Wounds
|
Paying attention to a sensory feature improves its perception and impairs that of others . Recent work has shown that a Normalization Model of Attention ( NMoA ) can account for a wide range of physiological findings and the influence of different attentional manipulations on visual performance . A key prediction of the NMoA is that attention to a visual feature like an orientation or a motion direction will increase the response of neurons preferring the attended feature ( response gain ) rather than increase the sensory input strength of the attended stimulus ( input gain ) . This effect of feature-based attention on neuronal responses should translate to similar patterns of improvement in behavioral performance , with psychometric functions showing response gain rather than input gain when attention is directed to the task-relevant feature . In contrast , we report here that when human subjects are cued to attend to one of two motion directions in a transparent motion display , attentional effects manifest as a combination of input and response gain . Further , the impact on input gain is greater when attention is directed towards a narrow range of motion directions than when it is directed towards a broad range . These results are captured by an extended NMoA , which either includes a stimulus-independent attentional contribution to normalization or utilizes direction-tuned normalization . The proposed extensions are consistent with the feature-similarity gain model of attention and the attentional modulation in extrastriate area MT , where neuronal responses are enhanced and suppressed by attention to preferred and non-preferred motion directions respectively .
Attention to visual features like a specific orientation or motion direction has been shown to enhance visual responses to the attended feature across visual cortex in both monkey neurophysiology [1] and human fMRI data [2–4] . Prior studies have reported that feature-based attention enhances responses in neurons tuned to the attended feature [5 , 6] , privileges responses to the attended feature under competitive conditions [7] and induces shifts of the preferred feature [8] . Similarly , visual attention to a particular spatial location affects neuronal responses and improves perceptual performance at the attended location [reviewed in 9] . In particular , attention has been shown to enhance neuronal responses by increasing the effective sensory input strength ( in our task: coherence gain: Fig 1A ) and/or by scaling the responses of the neuron ( response gain: Fig 1B ) [5 , 10–15] . The Normalization Model of Attention [NMoA: 9] attempts to capture this variety of attentional effects in a single model . It proposes that attention multiplicatively scales the driving input to a neuronal population , and the response to this driving input of each individual neuron in the population is divisively normalized by the responses of all the neurons in the normalizing pool . Depending on the size of the visual stimulus and the spread of visual attention , the relative effects of sensory stimulation and visual attention on the individual neuron and the normalizing pool differ , leading to input-gain and/or response-gain effects that reproduce many of the effects of spatial attention on neuronal responses [9 , 16] . Further , fMRI measurements of the spatial spread of visual attention in human subjects provide support for this critical assumption of the NMoA by verifying the model’s predictions regarding the influence of the spatial spread of visual attention on behavioral performance [17] , or voxel-averaged neurometric functions [18] . The NMoA also captures some of the reported effects of feature-based attention on neuronal responses [9] , using the same underlying mechanism of attentional scaling of sensory responses . Importantly , the NMoA predicts that , assuming biologically plausible parameters ( see Materials and Methods ) [19] , attention to a visual feature will impact neuronal responses mainly by increasing the effective response of neurons tuned to the attended feature ( response gain ) , rather than by increasing the sensory input strength of the attended stimulus ( input gain ) . This implies , given a quasi-linear linking-model relating neuronal responses to behavioral output [20] , that attention to a visual feature will not produce input-gain effects , but only response-gain effects on psychometric functions . Herrmann et al . [19] confirmed this prediction when they observed only response gain effects in an experiment where human subjects paid attention to either narrow or broad ranges of orientation . In contrast , we report here that when human subjects are cued to attend to one of two motion directions in a transparent motion display , attentional effects manifest as a combination of input gain ( in our task “coherence gain” ) and response gain . Further , different from conclusions drawn by Herrmann et al . [19] , we observed a larger impact on input gain for a narrow focus of attention in feature space than for a broad focus , while the observed response gain effect was not significantly different between conditions . These results require either a revision of the assumptions linking neuronal activity to behavior , or extensions of the NMoA that include direction-tuned influences on the normalization pool . Since given the assumptions of the linking model , psychophysical performance can be used to estimate neuronal responses [20] as well as to deduce models of divisive normalization [21] , we propose and compare two possible extensions to the NMoA , introducing either coherence-dependent or coherence-independent direction-tuned normalization . The extended normalization models are consistent with the feature-similarity gain model of attention [5] and the attentional modulation in extrastriate cortical area MT , where neuronal responses are enhanced and suppressed by attention to preferred and non-preferred motion directions respectively [6] .
To validate whether our cueing paradigm was effective in causing differential attentional deployments , we computed each subject’s mean performance across coherences , separately for each subject , attentional condition ( valid/invalid ) and cue type ( narrow/broad ) . We then performed four pair-wise comparisons ( Bonferroni corrected α = 0 . 0125 , paired t-tests , n = 6 subjects ) . Fig 3 shows the average of these coherence-averaged performances across subjects . For both attentional conditions , subjects performed significantly better when the cue was valid than when it was invalid ( narrow focus: mean Δd′ = 0 . 958 , p<0 . 001 , broad focus: mean Δd′ = 0 . 358 , p = 0 . 006 ) . Further , the performance for the validly cued direction was significantly better in the narrow focus condition compared to the broad focus condition ( mean Δd′ = 0 . 513 , p = 0 . 003 ) . The performance in the invalidly cued direction was not significantly different between the two attentional conditions ( mean Δd′ = −0 . 087 , p = 0 . 24 ) . For the statistical tests performed above , we repeated all comparisons with paired , two-sided Wilcoxon signed rank tests . This did not qualitatively change our results ( i . e . all statistically significant results remained significant and all non-significant results remained non-significant ) . The core aim of our study was to determine whether feature-based attention enhances performance by coherence or response gain and match our findings to the predictions of the NMoA . This was done by determining each subject’s coherence response function in each of our four task constellations by fitting Naka-Rushton equations ( Fig 4 ) with a shared slope parameter for the four conditions . The task was tailored individually to each subject ( see Materials and Methods section ) leading to comparable performances across coherences and to comparable model results across subjects . Indeed , performing pairwise t-tests on R2-values obtained for each subject and attentional condition , we did not observe significant differences in the goodness of fits for the four task conditions . Mean R2-values ( for 6 individually fitted subjects ) were 0 . 98 ( narrow-valid ) , 0 . 94 ( narrow-invalid ) , 0 . 98 ( broad-valid ) and 0 . 91 ( broad-invalid ) . We then compared the fitted Naka-Rushton coefficients for validly and invalidly cued trials , to test if attention induced a reduction in c50 and/or an increase in d′max . A decrease in c50 indicates an increase in coherence gain and an increase in d′max indicates an increase in response gain . We performed four pair-wise comparisons ( Bonferroni corrected α = 0 . 0125 , paired , one-tailed t-tests , n = 6 subjects , we also performed this analysis with paired , two-tailed t-tests , which did not change our conclusions ) . For the narrow focus condition ( Fig 4A ) , we find a significant cue-induced increase in coherence gain ( mean Δc50 = −0 . 179 , p = 0 . 002 , Fig 5A ) as well as in response gain ( mean Δd′max = 0 . 895 , p = 0 . 001 , Fig 5B ) . In the broad focus condition ( Fig 4B ) , the response gain enhancement is of similar magnitude and also significant ( mean Δd′max = 0 . 628 , p = 0 . 004 , Fig 5B ) while the coherence gain enhancement is much smaller and narrowly misses significance ( mean Δc50 = −0 . 062 , p = 0 . 047 , evaluated at α = 0 . 0125 , Fig 5A ) . These effects ( averaged across subjects ) are also evident in single subjects ( Fig 4 and Fig 5 ) . As plotting performance as d′ might amplify differences at high coherences , we also performed the same analysis based on the proportion of correct responses . This did not change the pattern of results ( i . e . response gain in the broad focus condition and a combination of coherence and response gain in the narrow focus condition ) . Next , we tested whether the magnitude of coherence ( c50 ) and response gain ( d′max ) changes with attentional condition ( i . e . with an increasing width of the feature-based attentional focus ) . We calculated a modulation index MIζ ( ( a − b ) / ( a + b ) , see Materials and Methods section ) for each coefficient-condition pair and then performed paired comparisons of the distribution of indices across attentional conditions . We find that the magnitude of coherence gain is significantly different between attentional conditions ( mean ΔMIc50=0 . 293≜82 . 9% , p = 0 . 007 , paired t-test ) , while there is no significant change in response gain ( mean ΔMId′max=0 . 033≜6 . 8% , p = 0 . 136 , paired t-test , Bonferroni corrected α = 0 . 025 ) . For these statistical tests , we repeated all comparisons with paired , two-sided Wilcoxon signed rank tests . This did not qualitatively change our results . We further addressed a potentially confounding ceiling effect of performances at high coherences by repeating the above analysis , leaving out the two highest coherences ( i . e . the highest performances we measured in our task ) of the valid condition in narrow focus trials , thereby disregarding data points that might have been affected by a ceiling effect of performance . With this reduced dataset , the increase in response gain narrowly misses significance in the narrow focus condition , however , a coherence gain change was still highly significant . The narrow focus cue did not signal the precise direction of the sample stimuli , but rather indicated that the relevant sample was likely to occur within a range of ±10 degree around the cued direction ( heading of the arrow ) . Nonetheless , we tested whether subjects used the cued direction as sample and simply ignored the subsequently presented sample direction . If this were true , direction discrimination performance should increase once the test direction was far off from the cued direction . Fig 6A shows the performance across coherences for three groups of trials that differ in how far off the cued direction the test direction occurred . Groups were defined individually for each subject based on his/her individual direction change magnitude ( see Materials and Methods ) and we divided the possible range of absolute cue-test differences into three evenly spaced parts ( close , medium and far ) . Since upcoming invalidly cued directions could also be inferred from the cue ( since the uncued direction range centered ±135 degrees from the cued direction ) , we were able to define the same three groups for invalidly cued trials . For each group we find significant effects of cue validity ( paired t-tests , p<0 . 001 , p = 0 . 002 , p<0 . 001 for close , medium and far , respectively ) while pairwise comparisons indicated that none of the three groups of validly cued trials was significantly different from the others . The same was true for the invalidly cued trials ( all p>0 . 027 , Bonferroni corrected α = 0 . 0083 , n = 6 comparisons ) . We thus find no evidence pointing towards subjects using the cue direction ( rather than the sample direction ) as a reference for the direction discrimination task in the narrow focus condition . We also tested whether sample presentations occurring far from the cued direction resulted in improved task performance . For this purpose trial groups were defined as sample directions close ( 0–2 degrees ) , medium ( 3–6 degrees ) , and far ( 7–10 degrees ) from the cued direction ( or the inferred uncued direction ) . Fig 6B shows the performance across coherences for those three trial groups . Similar to the trial grouping by sample-test difference , we find significant effects of cue validity ( paired t-tests , p = 0 . 001 , p<0 . 001 , p = 0 . 001 , for close , medium and far , respectively ) . Again , no pairwise comparison between groups was significant for either valid or invalid trials ( all p>0 . 02 , Bonferroni corrected α = 0 . 0083 , n = 6 comparisons ) . This suggests that in the narrow focus condition , the featural extent of attention covered at least a range of 20 degrees , centered on the attentional cue , which we also assumed in all model simulations . Our experimental results reveal a mixture of coherence and response gain enhancements when attention is focused on a narrow range of directions ( narrow focus condition ) , and a pure response-gain enhancement when attention is focused on a broad range of directions ( broad focus condition ) . As pointed out by Herrmann et al . [17 , 19] , a change in behavioral performance will mimic the underlying change in neuronal response functions , and therefore only a pure response gain for attention to motion directions will be visible in the neurometric function [20] . Further , even if any coherence gain effects were to arise , they would be found in the broad focus condition , which is the opposite of what our empirical data show . The intuition behind these statements has been presented in detail by Herrmann et al . [19] as well as Reynolds and Heeger [9] , but we summarize it briefly here: The NMoA computes the response of an arbitrary single neuron to a given set of stimuli as: R ( x , θ;c ) =Ai ( x , θ ) E ( x , θ;cn ) S ( x , θ;c ) +σn ( 1 ) where R ( x , θ;c ) is the response of a neuron with its receptive field centered at x and its feature tuning centered at θ , receiving stimulus input with contrast c . Ai ( x , θ ) E ( x , θ;cn ) is a term composed of the net excitatory input drive to the neuron E ( x , θ;cn ) scaled by the attentional gain Ai ( x , θ ) ≥ 1 , which varies with cue validity and attentional condition ( i . e . narrow or broad focus ) . Further , E ( x , θ;cn ) also depends on the stimulus contrast raised to an exponent ( cn ) while both E ( x , θ;cn ) and Ai ( x , θ ) depend on the similarity of the neuron’s receptive field and tuning properties with the driving stimulus and the attentional focus , respectively . S ( x , θ;c ) is the effect of the normalizing pool and represents the excitatory drive convolved by the suppressive surround: S ( x , θ;c ) =s ( x , θ ) *[Ai ( x , θ ) E ( x , θ;cn ) ] ( 2 ) where s ( x , θ ) is the suppressive filter ( defining the spatial and feature tuning of the surround ) and * indicates a convolution . For the transparent motion stimuli with two component motion directions that we used , the response of one neuron with preferred direction centered at one of the component directions ( from Eq 1 ) can be simplified ( without attention ) as: R ( c ) ≈αcS+σ ( 3 ) with α as the ( constant ) gain of the neuron receiving it’s preferred input with contrast c and S representing the net normalizing effect of the neurons in the population . S is regulated by the width of s ( x , θ ) ( see Eq 2 ) . When s ( x , θ ) is narrow ( strongly tuned normalization ) , attention ( γ ) acts equally on the driving input and the normalizing factor S and this leads to a coherence-gain effect ( Reynolds and Heeger 2009 ) : R ( c ) ≈γαcγS+σ=αcS+σγ ( 4 ) More explicitly , this happens because the normalizing pool is dominated by the inputs that excite the neuron and attention to the non-preferred feature is essentially invisible to the neuron since it lies outside both the excitatory and suppressive filters . In contrast , when s ( x , θ ) is broad , the impact of attention on the denominator S + σ is minimal ( even if the attentional spread is broad ) since the normalizing pool includes almost equal contributions from the neurons centered at the attended and unattended directions . Under these conditions , R ( c ) ≈γαcS+σ ( 5 ) which represents a response gain for the validly cued condition compared to the invalidly cued one . As a result , for the NMoA to predict a coherence-gain effect of attention , the normalizing pool ( or suppressive surround ) would have to be so narrow ( see below ) as to be physiologically implausible . Further , since the coherence-gain effect is facilitated when attention has a greater impact on the normalizing pool ( by acting more broadly ) , it is the broad focus condition that should show a stronger coherence-gain effect of attention . We confirm these statements by explicitly fitting the NMoA to our data . Free parameters , shared among attentional conditions , were the gain of attention ( Ai ) , separately optimized for narrow and broad conditions , the normalization constant σ , the exponent n and a scaling parameter to linearly scale simulated values to d′ ( for the values of the fixed parameters , see Materials and Methods section ) . The best fitting NMoA model shows a clear lack of fit to the empirical data ( Fig 7 ) , especially in the narrow focus condition , which is expected because that is where the coherence-gain effects manifest . The NMoA model’s best fit resembles a response gain in both attentional conditions , as expected . The observed lack of fit is not a result of our chosen fixed parameters: varying all but one of those parameters over a large range did not change our conclusions . The only critical parameter , as mentioned above , is the width of the suppressive filter in the feature dimension . We therefore redid the fits , but with the featural width of the suppressive filter as an additional free parameter ( NMoA free model ) . This resulted in an optimal , yet biologically implausible , inhibitory tuning width of σ = 12 . 3 degrees and a model producing clear effects of coherence gain in both attentional conditions ( Fig 7 ) . This model accounts for the reduction of coherence gain in the broad-focus condition by proposing that the broader width of the attentional field is accompanied by a reduced attentional gain . While this is not an unreasonable assumption , it compromises the ability of the model to account for the observed response-gain changes , especially in the broad-focus condition ( Fig 7B ) . Thus , even if the original NMoA is allowed to take on biologically implausible parameters , it still does not capture our data fully . Since the original NMoA does not capture our observed effects of feature-based attention , we attempted to extend the NMoA in the simplest , yet most plausible manner in order to do so . The empirical data indicate that the coherence-gain effect of feature-based attention emerges for the validly-cued feature and is greater in the narrow focus condition . One way to incorporate a coherence-gain effect is to postulate that in addition to enhancing the input drive to the attended feature , feature-based attention reduces the coherence-independent normalization term σn ( NMoA+ciN model ) and that this reduction is greater when attention is more focused ( as in the narrow focus condition ) . This reduction is independent of stimulus strength ( coherence ) and direction , but tuned to the attended direction such that attention to a particular motion direction reduces the normalizing effect on neurons tuned to that direction and potentially enhances the normalizing effect on neurons tuned to far-away directions . In other words , Eq 1 can be rewritten in an extended form as: R ( x , θ;c ) =Ai ( x , θ ) E ( x , θ;cn ) S ( x , θ;c ) +σnN ( θ ) ( 6 ) where 1≤N ( θ ) represents the direction-tuned effect of attention that is maximal for motion directions close to the attended feature . Another way to incorporate a coherence-gain effect is to unify the NMoA with models utilizing previously proposed ideas of neuronal self-normalization [e . g . 22 , 23] . Here , each neuron is normalized not only by its suppressive surround , but also by its own net-excitatory input . Such a coherence-dependent extension of the NMoA ( NMoA+cdN model ) can be written as: R ( x , θ;c ) =Ai ( x , θ ) E ( x , θ;cn ) N*Ai ( x , θ ) E ( x , θ;cn ) + ( 1−N ) *S ( x , θ;c ) +σn ( 7 ) where 0≤N≤1 is a single free parameter determining the balance between pure self-normalization ( N = 1 ) , predicting only coherence-gain , and the original NMoA ( N = 0 ) , predicting mainly response gain . We examine the potential physiological bases of both extended versions of the NMoA in the Discussion section . In terms of capturing the coherence-gain effects of attention , both models effectively capture both the response-gain and coherence-gain effects evident in our empirical data ( Table 1 and Fig 8 ) . We fit both extended NMoAs ( with one and two additional free parameters for the NMoA+cdN and NMoA+ciN model , respectively ) and compared them to the previously computed best fits from the original NMoAs ( fixed and free suppressive width , Fig 7 ) . Table 1 summarizes the results . Both extensions fit the data significantly better than the original NMoA ( F = 59 . 29 , p<0 . 001 , between NMoA and NMoA+cdN; F = 33 . 20 , p<0 . 001 , between NMoA and NMoA+ciN ) . Compared to the NMoA free model , only the NMoA+ciN model shows a significant advantage ( F = 0 . 98 , p = 0 . 56 , between NMoA free and NMoA+cdN; F = 8 . 60 , p = 0 . 004 , between NMoA free and NMoA+ciN ) . However , AIC as well as BIC measures indicate both extended NMoAs as superior to the original NMoAs . Between extended models , we find that the NMoA+ciN model performs marginally better than the NMoA+cdN model ( F = 5 . 23 , p = 0 . 024 ) with both lower AIC and BIC metrics for the NMoA+ciN model , confirming that the use of one extra parameter was justified and the model with a coherence-independent influence of attention on normalization described the data better than the model incorporating neuronal self-normalization .
The Normalization Model of Attention [9] has become the central model for capturing the known variety of attentional effects on neuronal responses , fMRI signals and behavioral performance . While the NMoA is powerful enough to explain a wide range of response patterns under physiologically plausible assumptions ( see Materials and Methods ) , it is also limited in flexibility and cannot predict certain patterns of responses , such as a reduction of input gain , but not response gain , caused only by a widening of the attentional focus . Since many assumptions underlying the NMoA’s parameters are not easily verified , such predictions of “impossible results” are critical because they allow the model to be stringently tested against empirical data . Here , we report that human subjects show behavioral performance patterns that go against a prediction of the NMoA and suggest and compare two simple and testable extensions to the NMoA that can account for the findings . As pointed out by Herrmann et al . [19] , the NMoA predicts that under biologically plausible parameter settings , attention to a visual feature like orientation or motion direction will only produce response-gain effects in neuronal response functions . Given that changes in the neuronal representation are assumed to scale quasi-lineary to behavioral performance [20 , 21] , these effects imply that similarly , only response-gain effects will be found when comparing psychometric functions measuring performance on tasks involving attended and unattended features . Herrmann et al . [19] went on to confirm this prediction by showing only response gain effects in psychometric functions when subjects paid attention to either narrow or broad ranges of orientation . Here , we built on this work by measuring the performance of human subjects on a task requiring them to discriminate a direction change in one of the two directions of a transparent motion display . Performance increased with motion coherence and was greater for validly cued stimuli . However , in contrast to Herrmann et al . ’s [19] results for attention directed to orientations , we found that attentional effects manifest as a combination of input gain and response gain on the psychometric function . Critically , when we compared the effects of attention directed towards either a narrow or broad range of motion directions , we found a significant decrease of input gain , but not response gain , for the broad focus , which cannot be readily accounted for by the original formulation of the NMoA . Our results using a motion direction discrimination task differ from those of Herrmann et al . ’s [19] task using orientation discrimination , despite the fact that the two tasks are conceptually very similar . One difference is that we varied coherence rather than contrast to manipulate signal strength in order to obtain a sufficiently large dynamic range . Currently , there is only limited evidence describing the effect of coherence changes on neurometric functions . Available results indicate that , at least for non-transparent motion patterns , the coherence-response function in MT is much more linear than the contrast-response function [24–26] . Sigmoidal coherence-response functions have also been reported in macaque MT [27] . It is not obvious why these differences between the coherence and contrast-response functions should cause the difference in our results . Our results show that adding either a coherence-independent contribution of attention to normalization or a coherence-dependent mechanism of self-normalization to the NMoA is sufficient to fully account for our data . This points to potential differences in the attentional contribution to normalization between our results and Herrmann et al . [19] . Further research is needed to determine how different stimulus properties and task demands might lead to different amounts of stimulus-dependent and stimulus-independent feature-based attentional contributions to neuronal normalization . We suggest two possible extensions of the NMoA both including direction-tuned influences on the normalization pool . The first model ( NMoA+ciN ) implements a coherence-independent , attentional contribution to normalization . Here , attention not only modulates the input drive to a neuronal population , but also reduces the impact of the normalization on the responses of neurons tuned for the attended direction . Further , the data indicate that such a tuned normalizing effect of attention would have to be greater when attention is more narrowly focused than when it is broadly distributed . To implement such a specific rescaling of the coherence-independent normalizing input in the brain , we suggest that since the NMoA can be considered a steady-state version of an unspecified network model with mutual competition , a stimulus at the preferred direction of the neuron could suppress the local population that is tuned to non-preferred directions and thereby reduce their contribution to the normalizing pool . Alternatively , we propose in the second model ( NMoA+cdN ) that each neuron preferentially weights its own contribution to the normalization pool ( self-normalization ) in comparison to the contribution of all suppressive neurons . Such a mechanism was previously shown to be a vital component in a model capturing the response properties of direction-selective neurons in extrastriate cortex [22] . The tuned normalization in another recent report [23] is also conceptually similar: here , the authors showed that MT neuronal responses to a pair of stimuli within the receptive field ( one moving in the preferred direction and the other in the anti-preferred direction ) were well explained by direction-tuned divisive normalization . The majority of neurons in their data showed a greater normalizing influence of the preferred stimulus . We show here that extending the NMoA with an explicit tuned-normalization component also captures our results in an attention task , despite the fact that this coherence-dependent mechanism is independent of the spread of attention . However , the difference between the two extensions is significant and the NMoA+cdN model described the data worse than the NMoA+ciN model . The proposed NMoA+ciN model modifies the normalization mechanism to include a reduction by feature-based attention of the normalizing influence for neurons tuned to the attended direction . There are a variety of ways in which this modification could be implemented . For example , if feature-based attention suppresses the responses of neurons tuned to non-preferred directions , their contribution to the normalization pool could be reduced thereby reducing the coherence gain for neurons tuned to the attended direction ( but increasing it for neurons tuned to the unattended direction , where the normalization pool will be enhanced ) . Alternatively , feature-based attention may enhance both the "stimulus drive" as well as the "normalization" for neurons tuned to the attended direction , and this effect may manifest as coherence gain . Importantly , here the direction selectivity of the normalization pool is not critical , but instead , attention has a selective effect on neurons tuned to the attended direction [12] . Thus , the mechanism works even if the normalization pool is untuned , but critically , it may also work when the normalization pool is tuned . In a related framework , Boynton [28] proposed a normalization model with a stimulus independent contribution of attention to the normalization pool . This untuned normalization can account for attentional effects of input gain when attention is directed inside versus outside of a neuron’s receptive field . For non-spatial forms of attention , as described here , a feature-tuned input to normalization is necessary since attention does not shift out of the receptive field . It should be pointed out , however , that the proposed extension with a coherence-independent , tuned input to normalization ( NMoA+ciN ) can similarly be applied to this or other previously proposed models of attentional normalization [16 , 21 , 28–30] . In addition to the extended normalization models considered above , one can imagine an important alternative to account for our empirical results . The hypothesized modifications all assume , that the behavioral effects of attention and its spread emerge from its effects on the neuronal representations of the stimulus ( i . e . the perceptual representation ) . However , attention may also act by modifying the decisional mechanism , for example , through enhanced weighting of the cued stimuli [31–38] . Specifically , the change in performance between validly and invalidly cued features could result from the differential weighting of inputs from the two motion directions , with greater weight given to the validly cued feature . With a lower weight to the unattended motion direction , the performance may only rise above chance once the coherence becomes sufficiently large . Similarly , the change in performance for validly cued motion directions between trials with focused or dispersed feature-based attention may be due to improved weighting of the same perceptual representation , rather than an effect of attention on the perceptual representation itself ( as we assume here ) . Differentiating between these two alternatives may require physiological recordings that examine the effects of feature-based attention under our conditions in the dorsal motion-processing pathway in order to measure the underlying neuronal coherence-response functions . Spatial attention has been shown to affect correlations within neuronal populations encoding visual features [39 , 40] and to reduce single-neuron variability [41 , 42] . Such effects can cause improvements in psychophysical performance even without increases in neuronal responses . The NMoA does not consider such attentional effects and thus aims to account for changes in psychophysical performance by changes in mean spiking activity . Consequently , we have assumed that the attentional modulation of psychophysical performance is independent of changes in correlations between neuronal firing of individual neurons . Additional experiments are needed to clarify to which degree feature-based attention causes changes in both neuronal correlations and neuronal variability and how those potential effects translate into changes in psychophysical performance . Attention to an anti-preferred motion direction suppresses the responses of MT neurons across the visual field in a multiplicative manner [5] . This finding inspired the feature-similarity gain model of attention which postulates that attending to a particular motion direction ( or more generally , visual feature ) enhances the responses of neurons tuned to the attended motion direction and suppresses the responses of neurons tuned to the opposite motion direction [6] . The NMoA can account for these findings by postulating that feature-based attention to the non-preferred direction increases its contrast or coherence-dependent contribution to the normalizing pool . Both of the proposed extensions to the NMoA do not compromise these previous predictions made by the NMoA , since they both contain the original model as a special case . However , the NMoA+ciN model has an additional mechanism whereby feature-based attention to the preferred direction has a coherence-independent “pure attentional” effect on the normalizing pool . This attentional influence can release a neuron from the suppressive effect of normalization when its preferred direction is attended . Measuring the extent to which these two effects contribute to the enhancing and suppressive effects of feature-based attention will require experiments specifically designed to tease apart these two effects . In summary , our results support and extend the popular NMoA with a modulatory mechanism specific to feature-based attention . This will allow the NMoA and similar models of attention and divisive normalization to cover an even wider set of conditions . As our extensions generate testable predictions , they are well suited to guide further research into the mechanisms and phenomenology of feature-based attention .
Eight subjects ( ages 18–27 years ) participated in the study , out of which 6 subjects ( 2 naive female , 3 naive male and 1 male lab member ) reached a sufficient performance level for analysis ( see section Data Analysis below ) . All subjects reported normal or corrected to normal vision . Prior to entering the main experiment four subjects participated in a pilot study to determine a suitable task timing . All naive participants received monetary compensation for each session . Subjects were verbally instructed about the task demands and received individual training before entering the main experiment ( see section Pre-Tests ) . All experiments were in accordance with institutional guidelines for experiments with humans and adhered to the principles of the Declaration of Helsinki . Each subject gave informed written consent prior to participating in the study . Stimuli were presented on a LCD screen ( SyncMaster 2233 , Samsung ) with a refresh rate of 120Hz and a background luminance of 20 cd/m2 . The experiment was controlled by an Apple computer ( MacPro 2010 ) running the open-source software MWorks version 0 . 5 ( mworks-project . org ) . Subjects were seated in a dimly lit room at a viewing distance of 57cm from the screen , their head resting on a chin-rest . A gamepad ( Precision , Logitech ) was used for recording responses , such that a button press with the right index finger indicated a clockwise decision , and the left index finger a counter-clockwise decision . Each experimental trial was started by pressing a button with the right thumb . For three subjects , eye position was recorded monocularly ( left eye ) using a video-based eye tracker ( IView X , SMI ) sampling at 250Hz . For the remaining three subjects , eye position was recorded binocularly with a sampling frequency of 500Hz using an Eyelink-1000 system ( SR Research ) . Both eye position systems were calibrated before each experimental session and the accuracy of the calibration confirmed by a custom calibration task . Fig 2 depicts the experimental paradigm . Subjects viewed moving random dot patterns ( RDPs ) through a stationary annulus-shaped virtual aperture with an inner diameter of 5 degrees and an outer diameter of 17 . 8 degrees of visual angle . The RDPs contained 4 dots/deg2 , moving on individual linear paths at a speed of 15 deg/s . Each dot had a diameter of 0 . 252 degrees and a luminance of 70 cd/m2 . Subjects had to maintain their gaze on a fixation point central to the RDP and to initiate each experimental trial by a thumb-button press . Then an attentional cue was presented ( see section "Attentional Cues" ) for 500ms on top of the fixation point . Following the cue and a 800ms delay , a RDP was displayed for 650ms . This first presentation of the RDP contained two superimposed groups of coherently moving dots ( ‘direction components’ ) , as well as an additional number of randomly moving dots . The two motion directions of this transparent motion display were always 135±20 degrees apart , with each direction being sampled randomly from a ±10 degree range around a reference direction . Reference directions were +45 , 0 and -45 degrees from straight left or rightward motion . The presentation of this first RDP was followed by a short delay of 100ms with only the fixation point present on the screen . Then the second RDP was displayed for 400ms , with a slightly rotated version of one of the two previously shown motion directions , as well as the same proportion of noise dots as in the first RDP . Subjects had to indicate whether the single motion direction of the second RDP was rotated clockwise or counter-clockwise relative to the closest motion direction of the first RDP ( 2 alternative-forced choice , Fig 2 ) . Subjects received auditory feedback indicating correct or wrong judgments . The magnitude of the direction change was individually set for each subject to be the pooled just noticeable difference of all reference directions ( see section Pre-Tests ) . We varied the motion coherence on a trial-by-trial basis . Motion coherence was defined as the percentage of dots moving in signal directions . The remaining noise dots moved on linear paths in random directions . The coherence level was the same for both presentations of the RDP ( i . e . regardless of how many motion directions were presented ) . We used 6 levels of coherences ( 1 . 6% , 6 . 4% , 12 . 8% , 25 . 6% , 51 . 2% and 100% ) for each attentional condition . Throughout each session , all cue types and coherence levels were pseudo-randomly interleaved . One session consisted of 576 properly terminated trials , excluding fixation errors and erroneous early responses . Each subject participated in 5 sessions for a total of 2880 analyzed trials per subject . Trials in which eye-positions occurred outside a radius of 2 . 5 degrees around the fixation point , or eye blinks were considered fixation breaks . They caused trials to be aborted with an auditory feedback to the subjects . On average across all trials the subject’s eye positions during both stimulus presentations remained within a circular window with a radius of less than 0 . 6 degrees . Previous studies aimed at developing or testing the NMoA have used spatially separated target and distractor stimuli , which could have been selected by spatial attention . We used a transparent motion display containing two spatially overlapping moving RDPs , leaving feature-based attentional mechanisms as the sole selection mechanism for behavioral enhancement . Two types of cues were used to direct subjects’ attention to one of the two motion directions of the transparent motion display . The narrow focus cue was a single arrow pointing in one of the six reference directions , indicating that the relevant motion signal of the first stimulus presentation was likely to occur within a range of ±10 degrees around its heading . The broad focus cue consisted of three arrows , all pointing either towards the left or the right side , indicating that the relevant motion was likely to be right- or leftwards . Both cues were valid ( i . e . the relevant motion occurred within ±10 degrees of the narrow focus cue and towards the side of the broad focus cue ) in 75% of all trials and all subjects were verbally instructed and frequently reminded to also pay some attention to the uncued directions . The narrow focus cue was designed to enable subjects to direct their attention onto a narrow range ( ca . 20 degrees ) of possible target directions , while the broad focus cue was used to induce a much wider focus ( ca . 110 degrees ) of the feature-based attention field . In both cases , attention helped the subjects to preferentially focus on one of the two directions of the transparent motion stimulus for subsequent comparison with the single motion . The frequency of occurrence for the different types of cues was balanced between cue directions and cue types , such that no cue direction or cue type was overrepresented . We determined the influence of feature-based attention on psychophysical performance by comparing validly and invalidly cued trials . Pre-testing consisted of 2 to 6 sessions of 450 valid trials each . Pre-test trials were identical to regular trials , but contained no attentional cues . Furthermore , the coherence level of all stimuli was set to 51 . 2% . To measure each subject’s individual just noticeable difference ( JND ) , we varied the direction change magnitude in 15 discrete steps from -14 to 14 degrees . We then fitted a psychometric function ( cumulative Gaussian ) for each subject and each reference direction . Subjects started the main experiment once they reached a comparable performance for all six reference directions , with little to no bias in their discrimination thresholds . The subject JND was defined as the slope of the cumulative normal fit of the performance pooled over all reference directions . Subjects were trained to perform the pre-task until they reached a JND smaller than 16 degrees in one complete session of testing , or until they aborted the experiment . Altogether , 23 subjects entered the pre-testing phase , out of which 8 subjects continued to the main experiment . Subjects aborting the experiment mostly reported that they found the task too demanding to commit to further training or testing . For subjects reaching the criterion , their JND from the last session of pre-testing was used throughout the main experiment ( mean JND = 12 . 86 , standard-deviation = 1 . 94 ) . To test whether the two types of attentional cues led to measurable attentional effects , we compared each subject’s mean performance over all levels of coherences between both attentional conditions . We calculated performance as d′ = zscore ( pCWcorrect ) − zscore ( pCCWfailure ) , where ‘pCWcorrect’ is defined as the proportion of clockwise responses to clockwise changes , and ‘pCCWfailure’ as the proportion of clockwise responses to counter-clockwise changes . Using paired t-tests we determined whether performance differed between trials with narrow and broad focus cues and confirmed that attention was deployed in line with each cue type , as indicated by a significant difference between validly and invalidly cued trials . In order to determine whether attention affected performance by response or coherence gain we investigated separately for each attentional condition , how each subject’s performance changes with motion coherence . To obtain the coherence response function , we fitted a Naka-Rushton equation [43–45] d′ ( c ) =d′maxcncn+c50n to each experimental condition using a non-linear least-squares procedure . Using this equation , psychophysical performance d′ for each level of coherence c can be described by the asymptotic performance at high levels of coherence d′max , the coherence level at half asymptotic performance c50 and the slope of the function n . We tested with one-tailed , paired t-tests whether changes in c50 and d′max occurred from invalidly to validly cued trials for each attentional condition . Significant increases in d′max represent response gain effects and significant decreases in c50 represent coherence gain effects . The slopes of the corresponding coherence response functions for each attentional condition were constrained to be equal in all four fits per subject to minimize the number of free parameters . We validated this choice by comparing this reduced model ( with a single exponent per subject ) to those with two exponents per subject ( one for each attentional condition ) and to those with four exponents per subject ( one for each attentional condition and cue validity ) . The reduced model with a single exponent per subject produced almost identical fits and was clearly preferred ( due to its lower number of parameters ) by AIC and BIC measures . We evaluated further-reduced models with shared parameters ( d′max or c50 ) either across or within attentional conditions , but found that no simpler model was superior to the one described above . A robust fit of the coherence response functions requires that the asymptotic performance saturates at high levels of coherence . We therefore excluded two subjects with performance increases of Δd′ ≥ 1 between the two highest coherence levels , leaving a total of 6 subjects for the final analysis . To determine the coherence gain and response gain changes between attentional conditions , we computed a modulation index for each of the gain enhancements: MIζ=ζvalid−ζinvalidζvalid+ζinvalid where ζ corresponds to one of the two fitted coefficients c50 or d′max . We calculated the differences in modulation magnitude between conditions and tested with paired t-tests if the effect sizes of coherence and response gain varied significantly between the two attentional conditions . All statistical tests were Bonferroni corrected for multiple comparisons . Data analysis was done using custom scripts in Matlab R2014a ( MathWorks ) . We used the Palamedes routines [46] for fitting psychometric functions and the Matlab Curve Fitting toolbox ( MathWorks ) for the non-linear fitting . To simulate our empirical data with the NMoA , we used custom Matlab scripts , based on the code of Reynolds and Heeger [9] . We changed the original code to use a circular von Mises distribution for both the stimulation and the attention fields’ theta dimension . Therefore we express the width of the feature-attention spotlight in terms of parameter κ , which is the concentration of the distribution around it’s mean ( 1/κ is roughly equivalent to σ2 of a gaussian ) . We confirmed that this modified model produces similar results to the original NMoA by comparing our results with the outcome of the Matlab scripts available on the authors’ website . We modeled our empirical results by defining a stimulus that is infinite in space , since no spatial position inside the annulus carried more relevant signal than any other and thus spatial attention could not have impacted psychophysical performance . Consequently we assumed that for modeling purposes , spatial attention was evenly distributed across all spatial locations . The two directions of the transparent motion display were modeled as two narrow bands in the theta dimension , each with a concentration of κ = 33 , corresponding to roughly 10 degrees σ . The means of the two signals were 135 degrees apart from each other , corresponding to the mean difference in motion directions of the transparent motion display . Assuming a quasi-optimal attentional allocation according to the task design we then simulated an attentional field with either a narrow or a broad focus of feature-based attention . The exact choice of field width turned out to be not critical for the main finding ( see Results section for details ) . The narrow focus was an enhancement with a concentration ( angular extent ) of κ = 15 around one of the signals . The broad focus was centered on the same direction ( i . e . as if it were a horizontal movement ) , but enhanced a much broader range of directions around it ( κ = 0 . 5 , which corresponds roughly to 90 degrees σ ) . Our model MT population was defined to have Gaussian receptive fields with a spatial extend of σ = 5 degrees and a tuning width of σ = 37 degrees . The suppressive field was defined to have a spatial kernel width of σ = 20 degrees and a feature tuning width of σ = 180 degrees . The latter parameter was used since it is known that in motion selective area MT , surround tuning is present , but is generally very broad [47] . Overall , this biologically plausible set of parameters is very similar to the one used in previous simulations by Herrmann et al . [19] or Reynolds & Heeger [9] . We modeled increasing levels of coherence by increasing the value of the sensory input strength parameter c . In the NMoA , this essentially equates increases in coherence to increases in contrast . This choice ( also made by Jazayeri and Movshon [48] in a related context ) is supported by the physiological finding that MT units do not change their tuning for linear motion with changes in motion coherence [49] . In order to convert the modeled population activity into a prediction of behavioral performance , we assumed that task performance is dominated by the quality of decoding of the two motion directions of stimulus display 1 . Consequently , we selected two units of the simulated population with their tuning centered on the corresponding directions of stimulus display 1 ( out of which one was previously cued and thus in the focus of attention ) . We assumed that task performance on validly and invalidly cued trials is proportional to the values of the neurometric function for the attended and unattended unit respectively . A large value of the neurometric function translates to a greater signal-to-noise ratio for the neural representation and a better identification of the stimulus directions . Since the direction-difference between the sample and test directions was small , units tuned to the sample directions also responded strongly to test directions and received levels of attentional enhancement similar to units tuned to the test directions . Therefore , their neurometric functions would also be proportional to detection performance for presented test stimuli . In order to obtain the neurometric functions for relevant units , we repeated the simulation for varying values of c ( i . e . signal to noise ratios of the two bands in theta ) . Through appropriate rescaling with just one additional parameter , we converted the neuronal activity of the relevant unit ( depending on cue validity ) into psychophysical performance . Importantly , as shown by Pestilli et al . [20] , such a readout which equates attentional effects on neuronal response functions with those on behavioral psychometric functions ( after a rescaling ) leads to the same conclusions as those given by a more detailed implementation of an ideal likelihood-based observer [48] . Even when using this ideal observer to predict behavioral psychometric functions from the underlying modeled neuronal representation , the attentional effect on the behavioral psychometric function mimics the attentional effect on the underlying neuronal functions .
|
We report a pattern of feature-based attentional effects on human psychophysical performance , which cannot be accounted for by the Normalization Model of Attention using biologically plausible parameters . Specifically , this prominent model of attentional modulation predicts that attention to a visual feature like a specific motion direction will lead to a response gain in the input-response function , rather than the input gain that we actually observe . In our data , the input gain is greater when attention is directed towards a narrow range of motion directions , again contrary to the model’s prediction . We therefore propose two physiologically testable extensions of the model that include direction-tuned normalization mechanisms of attention . Both extensions account for our data without affecting the previously demonstrated successful performance of the NMoA .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"social",
"sciences",
"neuroscience",
"cognitive",
"psychology",
"computational",
"neuroscience",
"vision",
"neuronal",
"tuning",
"animal",
"cells",
"behavior",
"psychology",
"cellular",
"neuroscience",
"cell",
"biology",
"single",
"neuron",
"function",
"neurons",
"psychophysics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"sensory",
"perception",
"computational",
"biology",
"cognitive",
"science",
"attention"
] |
2016
|
An Extended Normalization Model of Attention Accounts for Feature-Based Attentional Enhancement of Both Response and Coherence Gain
|
A collective-risk social dilemma arises when a group must cooperate to reach a common target in order to avoid the risk of collective loss while each individual is tempted to free-ride on the contributions of others . In contrast to the prisoners' dilemma or public goods games , the collective-risk dilemma encompasses the risk that all individuals lose everything . These characteristics have potential relevance for dangerous climate change and other risky social dilemmas . Cooperation is costly to the individual and it only benefits all individuals if the common target is reached . An individual thus invests without guarantee that the investment is worthwhile for anyone . If there are several subsequent stages of investment , it is not clear when individuals should contribute . For example , they could invest early , thereby signaling their willingness to cooperate in the future , constantly invest their fair share , or wait and compensate missing contributions . To investigate the strategic behavior in such situations , we have simulated the evolutionary dynamics of such collective-risk dilemmas in a finite population . Contributions depend individually on the stage of the game and on the sum of contributions made so far . Every individual takes part in many games and successful behaviors spread in the population . It turns out that constant contributors , such as constant fair sharers , quickly lose out against those who initially do not contribute , but compensate this in later stages of the game . In particular for high risks , such late contributors are favored .
Cooperation , between selfish individuals in public goods games [1]–[8] , becomes particularly challenging when the prospects are uncertain and a critical number of cooperative acts is required . Investing in the prevention of climate change is in vain if too many other do not invest [9]–[13] . In this context , it may not only be important if we cooperate at all , but also when we cooperate . Motivated by the prospect of dangerous climate change , Milinski et al . have conducted a behavioral experiment to address such a situation [9] . The experiments were designed to capture a collective-risk social dilemma which arises when a group of individuals must cooperate to reach a common target in order to avoid the risk of collective loss . Subjects were distributed into groups of six players and given an initial endowment of money units ( in their case , each unit was worth 2 € ) . Over 10 rounds , each player could invest , or units into a common account . Preceding each investment decision , players were informed about the individual contributions in the previous round . At the end of the game subjects were allowed to keep their savings only if the common account contained at least half of the total endowment of the group; otherwise , all members lost all their savings with a certain risk probability . Milinski et al . found that when this risk is high , contributions increased overall . However , the majority of groups missed the target by a small margin , which is the worst possible outcome; investing nothing would lead to a higher expected payoff . The experiment of Milinski et al . has triggered numerous theoretical investigations [14]–[17] . The focus has been to use an evolutionary game in order to analyze the consequences of a target threshold , which represents a serious complication over the usual public goods games [8] , [17]–[21] . However , these studies considered only two behaviors , cooperation or defection , and assumed that individuals do not react to the contributions of their co-players over the course of the game . This means that effectively in these previous investigations the game was limited to a single round even though the full game consists of multiple rounds . However , a direct influence of co-players on individuals' decision emerges when there are several subsequent stages [22]–[24] of investment and it is not clear whether individuals should contribute in early or in late stages of the game . Herein , we explore the evolutionary dynamics of strategic behavior in such multi-round game by analyzing the timing of the contributions . With this method , we aim to understand the natural behavior in such kind of situations . This behavior is of particular relevance in the context of dangerous climate change , which has been modelled as a collective-risk dilemma [9] . Should we be pessimistic towards the prevention method used for climate change , especially when major industrial nations fail to fulfill their targets in CO reduction in time ? Or is this a natural behavior in such collective-risk dilemmas ? Under which circumstances would early contributions be favored ? In order to investigate strategic behavior in this game , we explore the general characteristics of such behavior through large scale computer simulations . We use evolutionary game dynamics [25]–[27] to infer which strategies are particularly stable in collective-risk games .
We employed an evolutionary game , in which success is measured by the average payoff over many collective-risk dilemmas . Such a collective-risk game is played among individuals selected at random from a well mixed population of size . An individual player commences each game with an initial endowment of , where is the total number of rounds played in a game . In each round , players simultaneously invest units into a common pool . The total investment of a player is . In our analysis , we focused on a six player game in which players can invest 0 , 1 or 2 units for ten rounds , as in [9] . We also discuss the consequences of relaxing these assumptions . The whole group collectively has to invest a target sum by the end of the game after rounds . If they succeed , they can keep what they have not invested . If they fail , they lose what they have not invested with probability and keep it with probability . Thus , a player obtains a payoff of when the target is reached and an average payoff when the target is missed . Note that the individual payoff is independent of the timing of the contributions – but this timing can be crucial for the interactions among the players . This collective-risk dilemma has a large strategy space and a large set of Nash equilibria . Each situation in which the group of players collectively contributes exactly and no player invests more than is a Nash equilibrium , irrespective of the distribution of contributions within the group . For example , for , half of the players could invest 2 units in each round and half of them nothing is a Nash equilibrium , despite being unfair . In this situation , the target is exactly met . If those who invest 2 units would invest less , the target would not be met . If those who invest nothing start contributing , these contributions would be in vain . In general , such deviations from the Nash equilibrium are disadvantageous for the individual in high risk situations . In addition , the situation in which no one contributes is a Nash equilibrium , because it takes more than one player to reach the target . A behavior can be defined from the individual contributions over the rounds . In our case , each player can choose between three actions in each round , thus there are different behaviors , increasing exponentially with . If behaviors are independent of the actions of others , we can collapse the whole dynamics into a single round game and identify strategies such as defectors ( someone who does not contribute , ) , fair sharers ( contributing half of the endowment , ) , altruists ( contributing everything they have , ) , or many others . However , when behaviors also depend on the actions of the other players , identifying the underlying strategies becomes much more challenging . In our case , the different behaviors are only based on the total amount that has been invested so far , a reasonable assumption in a context where it is difficult to monitor individual actions . Nevertheless , this assumption can lead to complex strategies and behaviors . A player's strategy determines how much to contribute in a given round , depending on the collective contributions so far . We assume that players invest more ( or less ) once the collective contributions have reached a certain amount . A player could aim to invest less when contributions are high , but it may also be reasonable to compensate the missing contributions of others . We defined a player's strategy based on a threshold and the contributions when the invested sum so far is above or below this threshold . For instance , a player could invest 2 in round if the total investment so far is above his threshold value and 1 otherwise . The contributions and thresholds can be different for each round , see Methods for a concrete example . This combination produces a large strategy space . Note that , an individual with a specific strategy ( defined by the thresholds and contributions ) can show a wide range of behaviors based on the common pool and hence on the strategies of other co-players . In evolutionary game dynamics , the payoff determines the fitness and thus more successful strategies spread in the population . In our setup , ‘evolution’ operates at the level of strategies while ‘selection’ operates at the behavioral level . Evolutionary game dynamics were simulated using a mutation-selection process in a population of finite size [27] , cf . Methods . The evolutionary game dynamics of strategic behavior depends crucially on the risk probability . As an illustration , Fig . 1 shows typical simulations for low risk ( ) and high risk ( ) , the parameter values analyzed in a behavioral experiment with students by Milinski et al . [9] . Within the first 200 generations , the average contributions and the average payoff values stabilize . As expected , for individuals do not contribute and the average payoff is of the initial endowment , cf . Fig . 1a . In contrast , for , individuals on average contribute half of their endowment ( ) , cf . Fig . 1b . In this case , the target is reached with a probability larger than 80% , leading to an average payoff substantially larger than . Note that the average payoff when the target is met , , is identical to the average payoff with zero contributions for . When , it is not worthwhile to contribute to the common account , because the expected payoff for not reaching the target is still higher than the payoff when the target is met and everyone contributes half of their endowment , . We find that , simulations for risk values up to lead to an average payoff of ( . Our simulations show that for the average payoff increases to values close to half of the initial endowment , which would be the optimal solution for high risks . This happens when the probability to meet the target reaches values much larger than 50% , see Fig . 2a . The probability to reach the target decays when there are more errors in strategy inheritance - they lead to changes in the contribution patterns which make it more difficult to evolve a solution for the game . We incorporate errors in our evolutionary process with probability , cf . Methods . Consequently , the average payoff decreases with increasing error probability . The diversity also increases for smaller intensity of selection . Increasing stabilizes the population faster and quenches the overall effects of . The dynamics of strategies can also be addressed on the behavioral level , which reflects the interaction of players and the corresponding strategic aspects . We use the contributions of individuals to differentiate between the different behaviors for games under various . A behavior with represents the classical defector , such players would always invest . E . g . in a four round game investing in each round , they would have as the corresponding behavior . The opposite behavior , an unconditional altruist , is represented by , which means the player contributes 2 in each round , e . g . in a four round game . A behavior with a represents any behavior where a player contributes half of the endowment over the rounds; there are many corresponding behaviors , e . g . in a four round game such a player could contribute a total of 4 units in 19 different patterns , 1 in each round ( 1111 ) , 1 in two rounds and 2 one round ( i . e . 0112 , 0121 , 0211 , 1012 , 1021 , 1102 , 1120 , 1201 , 1210 , 2011 , 2101 , 2110 ) , or 0 in half of the rounds ( i . e . 0022 , 0202 , 0220 , 2002 , 2020 , 2200 ) . In general , there are behaviors with , increasing rapidly with . Note that each of them – and any mixture of them – is a Nash equilibrium . For an efficient analysis we divided strategies into four behavioral categories , , , , and , see Fig . 2b . The behavior occurs at high frequencies for , while the behavior dominates for . Behaviors where occur for all at low frequencies , while over-contributors , , are also rare but only seen for very large . There is a single behavior associated with , however , there are many behaviors with . The increase in frequency of when could be attributed to any of them . Therefore , we divided the game into two halves and analyzed the contributions . It turns out that at least of the total contributions are made in the second half of the game for , Fig . 2a . Next , let us infer which behaviors are responsible for meeting the target when the risk is high . Interestingly , a single behavior dominates for all which can be described as a ‘fair rational’ behavior . The name indicates that these players invest their fair share , but also employ a reasoning related to backward induction for the strategic timing of their contributions . In this case , half of the endowment is contributed in the second half of the game and nothing is contributed in the first half of the game ( e . g . in a game with the dominating behavior can be represented as 0022 ) . Such behavior is consistent with the contribution increase observed in the second half of the game , Fig . 2a . We find this for different round numbers ranging from to , and a wide range of the other parameters ( see Fig . S1 ) . For instance , if we vary the maximum contribution permitted in each round , the same ‘fair rational’ behavior emerges , with contributions starting as late as possible ( see Fig . S2 ) . This indicates that the ‘fair rational’ behavior is preferred in such collective risk game when risk is high . We assessed the robustness of such behavior by initializing a homogenous population and analyzing the duration for which the behavior is maintained at a frequency greater than half of the initial population size . In Fig . 3 , we analyze five different behaviors: Non-contributors with , e . g . a 4 round game would have as the corresponding behavior , altruists with , i . e . , and three behaviors with : ( i ) the ‘fair rational’ , i . e . , ( ii ) fair naive , i . e . , and ( iii ) the reverse of the ‘fair rational’ , i . e . . Simulations show that as increases the stability of the ‘fair rational’ behavior improves . was most stable for all . For the stability of the behavior was similar to the ‘fair rational’ . When the ‘fair rational’ behavior is more stable than all other behaviors including the defecting behavior . The stability of the ‘fair rational’ behavior indicates that later contributions are favored for high risk , in line with our simulations of the mutation-selection balance . Our approach allows us to explore the impact of several aspects that have not yet been analyzed in a behavioral experiment . For a comprehensive analysis , we considered the effects of group size , interest on the common account , uncertainty in target , and continuously decreasing risk curves . First , we explored the impact of group size in such collective-risk game . When only few players have to coordinate their actions , a smaller strategy space has to be explored . In a game with , players do not invest for , for , players invest up to half of their endowment and at , more than half of the games meet the target . Investment still mainly occur in the second half of the game; behavior occurs at high frequencies for , while the behavior dominates for . Behaviors where occur for all at low frequencies , while over-contributors , , are also rare but again only seen for very large , see Fig . 4a–b . Furthermore , simulations show that when players are in smaller groups , contributions start at a lower risk value , compared to larger groups . For instance , for , contributions started for ( Fig . 4a ) , for contributions start at at ( Fig . 2a ) , and for , contributions started for ( Fig . S1a–d ) . Consequently , the payoffs increase to values above only for higher risk probability in larger groups . Second , we added an interest on the common account , such that early investments are more valuable . This only has an impact if the interest is high enough to replace a late contribution by a smaller , earlier contribution . For instance , simulations show that contributions begin to increase when increases above , when there is an interest of on the common account . When , the target is met with probability larger than 50% , in this case cf . Fig . 4c . We also observed that the behavior occurs at high frequencies for , while the behavior dominates for , see Fig . 4d . It is now possible to reach the target with such behavior . However , behaviors where occur for all at low frequencies . It turns out that ( unlike in the simulations without interest ) contributions are made in the first half of the game , Fig . 4c , this was consistent for different group sizes , cf . Fig . S1e–h . Interest also substantially increases the noise in the system; when interest was added to the common account individuals had an incentive to contribute early , however as a result invaders infiltrated and disrupted the stable equilibrium . Finally , we considered the effects of uncertainty in the target and smooth risk curves . If the target is not exactly known , it is substantially more difficult to evolve cooperation . Adding noise to the target causes the contributions to start at higher risks , but also causes a drastic decrease in the probability that the target reached . For example , without such noise and the target was reached with a 95% probability . But for a target subject to Gaussian noise with standard deviation of , the target was reached with only 80% probability , this was consistent for different group sizes , cf . Fig . S1i–l . Failure rate increased with increasing uncertainty in the target , for instance a target subject to Gaussian noise with standard deviation of dropped success to a 50% probability , Fig . 4e . Such uncertainty caused a change in behaviors , this is observed by a frequency increase in overcontributors ( ) and noncontributors ( ) and a decrease in the fair sharers ( ) , Fig . 4f . Despite the increase in failure probability , contributions reached half of the endowment when , Fig . 4e . We also considered a risk curve that is smooth instead of the step function , such that higher contributions continuously decrease the risk . Also in this case , the general picture does not change - late contributions are favored for sufficiently high risk .
The collective-risk dilemma is characterized by thresholds which capture risky collective-actions . Due to its potential relevance for dangerous climate change and other global crisis or risky social dilemmas , the general characteristics underlying such game structure are of crucial interest . Our model captures strategic elements in collective risk dilemmas by allowing individuals to interact and influence each other . We extracted a robust natural behavior for different risk levels . Our simulations of the collective-risk game unveiled a high abundance of a ‘fair rational’ strategy , such that the fair share is relinquished as late as possible . We vary the maximum contribution allowed , interest and uncertainty and analyze how all these factors influence the timing of contributions . We show that maximum payment dictates when contributions commence . Players procrastinate their contributions as much as possible . This implies that the maximum contribution possible ( or allowed ) per round determines the timing of contributions . Additionally , we show that interest to the common account can also affect the timing of contributions–individuals had an incentive to contribute early . This suggests that for time sensitive collective actions , incentives can be used to induce earlier contributions . We also show that uncertainty can cause a lack of coordination; simulations resulted in a decrease in success when the target was uncertain . Failure can also arise from an increase in group size or a decrease in risk probability . For larger group size the probability for successful cooperation decreases , as an individual's probability of being pivotal declines . However , simulations show that increasing risk probability quenched some of these uncertainties , and in turn , contributions increased . This suggests that chances of success increase when all the uncertainties are resolved . Moreover , it is essential to be informed about the maximum possible contribution , otherwise one may be too optimistic about the possibility to compensate in later stages . Finally , to understand the differences between the sequential game where individuals play in sequence [22] , [28]–[35] and the collective-risk game where individuals play simultaneously in a sequence of rounds [9] , [11] , [12] , we expanded the scope of our computational model by allowing for sequence allotment for individual players , see Fig . S3 . Our simulations reveal that a sequential game with 10 players has a lower efficiency in comparison to a 10 round collective-risk game for high risk probabilities . The collective-risk game requires coordination and , thus , a player needs to rely on deductive reasoning to ensure others are in agreement . A player should not invest if the chances are low that the other group members will invest sufficiently . Thus it is important to anticipate how others will behave . As the group number increases influencing – or predicting – the behavior of co-players becomes exceedingly difficult . However , deductive reasoning shows that the number of possible behaviors differ from round to round . If players wait and the pool remains empty , the number of strategies that allow to meet the target quickly drops . When nothing has been invested in the first half of the game , there is a single fair share behavior remaining , the ‘fair rational’ . By simply waiting , players are forced to play the ‘fair rational’ strategy or risk losing everything; such natural enforcement is most effective when risk is high . One rationale for the ‘fair rational’ behavior is that it can induce other players to contribute half of their endowment . Early contributions may seem more intuitive in the face of high risk . But they allow the invasion of other strategies and endanger the success of the game . Numerically , all fair share behaviors have the same payoff , but they diversify when considering invasions by deviating types . However , deviation from the ‘fair rational’ behavior is unforgiving and decreases the fitness of the deviating individuals , either collectively when the target is not met or individually when deviators have contributed too much . The ‘fair rational’ behavior leaves no room for conscious or erroneous deviation , and all co-players must contribute or risk consequences of failure for all . Players aiming to ‘play it safe’ by overcontributing or contributing early do not necessarily have a positive effect in collective-risk dilemmas , which can by their very nature lead to a detrimental outcome for all players . Thus , it is beneficial to have a strict behavior enforcing others to act alike , especially when stakes are high .
Each individual has a strategy composed of a threshold , and the contributions above and below the threshold for each round . The investment is thus determined by a player's strategy as well as the collective contributions so far . For instance , a player could invest 2 when the collective contributions are above ( or equal to ) his threshold and otherwise . We denote such a strategy in round by . A player that aims to compensate the missing contributions of others could instead have a strategy , such as - he would invest if the threshold is not met , but he would stop investing once the common pool is sufficiently filled . As an example , consider a game with two players and two rounds , and . The strategy of player one is . Player two has strategy . Since the common pool is empty in round 1 , , we have for player one , who thus invests . For player two , we have , which leads to an investment of . Now , in round we start with a common pool . Consequently , for player one – an investment of . Also for player two , we have – an investment of . As a result , the total investment after two rounds is and the target is met . Thus , player one obtains a payoff of and player two . Since payoff determines fitness , this means that the strategy of player two tends to spread . At the beginning of our simulations , all individuals have different random strategies , i . e . all contributions are , , or with the same probability and all thresholds are uniformly distributed between 0 and 1 . In one generation , such game are played , such that an individual on average plays games . The individual's payoff , , is calculated as the average payoff of all games played . At the end of a generation , the payoff is translated into a fitness value , where measures the intensity of selection [36] , [37] . Higher payoffs increase an individual's reproductive potential towards the next generation . The next generation is selected using the Wright-Fisher process where the individual's fitness is used to weigh the probability of choosing an individual for the new population [38]–[40] . Offspring inherits the strategy of the parent at the end of a generation ( games ) . We also incorporate errors in this process . Errors occur with a probability for the thresholds and the investments of each round independently . If they occur , errors in the threshold values add Gaussian noise with standard deviation to them . If an error in a contribution occurs , a random contribution is chosen , e . g . in the example above , in round 1 could be replaced by Once a new population is selected the process is repeated for multiple generations and the average of the dynamics is analyzed . To explore the stability of the different behaviors , a homogenous population was initiated using the same strategy for all individuals . In the simulation the population evolved under selection and mutation parameters . For each strategy we calculated the duration when the frequency dropped below half of initial population size , since this is a natural requirement for another strategy to take over ( all our strategies can be invaded in a finite population by neutral drift . So eventually , any strategy will be replaced due to mutation , selection and drift ) . Averages were computed over generations from realizations . Our simulations are written in C++ and were run on a 240 core Linux cluster . The computer code is available upon request .
|
The evolution of cooperation is a fascinating topic with a wide range of applications , from microbial evolution to global cooperation of humans in the context of climate change . Motivated by the prospect of dangerous climate change , behavioral experiments of a ‘collective-risk dilemma’ were conducted , where cooperation is in vain unless a threshold is met . This game requires multilateral efforts over several rounds in order to reach a known target and avoid collective loss . We have conducted large scale computer simulations to explore the evolutionary dynamics of strategic behavior in such collective-risk dilemmas . Individuals can react to the contributions of their co-players over the course of the game and adopt their own contributions . The timing of contributions to the public good is a very important issue for long-term problems such as climate change . In this context , it is imperative to know when individuals ( or countries ) would naturally contribute . We show that a specific behavior , late contributions , is favored , especially when risk is high . Collective-risk dilemmas can by their very nature lead to a detrimental outcome for all involved , and , thus it is crucial to understand the behavior that is expected in such a situation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"theoretical",
"biology",
"biology",
"evolutionary",
"biology"
] |
2012
|
Evolutionary Dynamics of Strategic Behavior in a Collective-Risk Dilemma
|
Developmental system drift is a likely mechanism for the origin of hybrid incompatibilities between closely related species . We examine here the detailed mechanistic basis of hybrid incompatibilities between two allopatric lineages , for a genotype-phenotype map of developmental system drift under stabilising selection , where an organismal phenotype is conserved , but the underlying molecular phenotypes and genotype can drift . This leads to number of emergent phenomenon not obtainable by modelling genotype or phenotype alone . Our results show that: 1 ) speciation is more rapid at smaller population sizes with a characteristic , Orr-like , power law , but at large population sizes slow , characterised by a sub-diffusive growth law; 2 ) the molecular phenotypes under weakest selection contribute to the earliest incompatibilities; and 3 ) pair-wise incompatibilities dominate over higher order , contrary to previous predictions that the latter should dominate . The population size effect we find is consistent with previous results on allopatric divergence of transcription factor-DNA binding , where smaller populations have common ancestors with a larger drift load because genetic drift favours phenotypes which have a larger number of genotypes ( higher sequence entropy ) over more fit phenotypes which have far fewer genotypes; this means less substitutions are required in either lineage before incompatibilities arise . Overall , our results indicate that biophysics and population size provide a much stronger constraint to speciation than suggested by previous models , and point to a general mechanistic principle of how incompatibilities arise the under stabilising selection for an organismal phenotype .
The genotype-phenotype map we use is a modification of the one described in [17] . The evolutionary task set for the gene regulation module is to turn an exponentially decaying morphogen gradient ( M ) across a field of cells in an embryo into a sharp step function profile of a downstream transcription factor T with its transition at the mid-point of the embryo , as shown in Fig 1 . This is accomplished by having the morphogen and an RNA Polymerase R bind to two adjacent non-overlapping binding sites in the cis-regulatory region ( C ) region of the transcription factor , the promoter P , and a single binding site B adjacent to it; transcription occurs whenever the polymerase binds to the promoter , although both proteins can bind to both binding sites dependent on their binding affinities . Binding to the regulatory region is cooperative due to stabilising interactions between the two proteins when bound at the two adjacent sites . The sequences of M and R at the DNA binding sites are represented by binary strings of length ℓpd = 10 . The corresponding DNA binding sites B and P are also represented by binary strings of the same length . Interactions between a pair of proteins are similarly represented by binary strings of length ℓpp = 5 . We assume an exponential morphogen concentration profile [M] ( x , α ) , as a function of the position of embryonic cells , x; the decay rate of the morphogen α is represented as a continuous variable , with a relative probability of mutation corresponding to an effective string of length ℓα = 10 bases . This results in a genome G , of total length ℓG = 60 . Protein-DNA and protein-protein binding strengths are determined by the number of mismatches between corresponding strings on the two interacting molecules , where for protein-DNA binding the cost of a mismatch is ϵpd = 2kBT and for protein-protein interactions ϵpp = 1kBT , where kBT is Boltzmann’s constant multiplied by room temperature ( 298K ) . We assume that there is a fixed concentration [R] of polymerase , in each cell . We then follow [28] and assume that the concentration of the transcription factor in a cell at position x ( [T] ( x ) ) is simply proportional to the probability of the polymerase being bound to the promoter , where this is calculated using standard methods of equilibrium statistical mechanics allowing for all configurations of protein species bound at these two binding sites , as well as none being bound ( see Methods for details ) . The fitness contribution F of the overall patterning phenotype ranges from 0 to κF depending on how well expression of the transcription factor is confined to the anterior half of the embryo , as shown in Fig 1 ( bottom left ) , where κF is a measure of the relative contribution of this trait to the fitness of the organism . We define a population-scaled fitness contribution 2NeκF , where Ne is the effective population size; for 2Ne κF < 1 the effects of selection are weak , and are conversely strong when 2NκF > 1 . We also assume that there is a boundary at F = F* , below which the organism is unviable . We simulate evolution as continuous time Markov process . After evolving a single population for a given number of generations , we form two replicates of the population that evolve independently , representing the process of allopatric speciation . At various time points following this imposed isolation , we consider the fitness and viability of various outcrossings between the two populations . A DMI occurs when the fitness contribution of a particular hybrid drops below F* .
The properties of a similar genotype-phenotype map have been previously explored [17] . An important property of this genotype-phenotype map is that only a single mechanism of patterning is found , in which the polymerase ( R ) binds with intermediate affinity to the promoter ( P ) but with high affinity to the morphogen ( M ) , while the morphogen binds to the morphogen binding site ( B ) only above a critical morphogen concentration . This results in a spatial switch once the morphogen falls below this concentration; evolution then fine tunes the relationship between the protein-DNA binding energies , the protein-protein binding energy and the steepness of the morphogen gradient α to turn off transcription at the mid-point of the embryo . Despite a single global solution there are many different combinations of the protein-DNA and protein-protein binding energies and α that give good patterning , and each of these correspond to many possible genotypes ( G ) . Of the different possible binding energies , we find that E M B , E R P , E ˜ R M ( binding energy of M to B , R to P , and R to M , respectively , calculated by the equivalent of Eq 3 in the Methods ) are under strong selection , whilst the other possible binding energies are essentially neutral with weak selective effects . At large population sizes it is found that the evolutionary dynamics exhibits what is known as quenched-disorder in statistical physics , where energy phenotypes that are less constrained take different random values between independent evolutionary runs with no further substitutions; this indicates an underlying roughness to the fitness landscape and that these weakly selected traits are trapped in a local optimum [17] . A key property determining the rate at which incompatibilities arise is the distribution of common ancestor phenotypes as a function of the population-scaled fitness contribution 2Ne κF , as shown in Fig 2 . For a given value of κF , we see that for large population sizes ( 2NκF ≫ 10 ) the distribution is what we would expect from conventional evolutionary theory on a fitness landscape with a fitness maximum . In contrast , as the population size is decreased , we find the distribution shifts to lower fitness values to the point when selection is weak ( 2NκF ≤ 1 ) the distribution is poised at the inviability boundary . This effect arises due to genetic drift at low population sizes pushing populations towards marginally fit phenotypes that correspond to the largest number of genotypes , that is , with the largest sequence entropy . Our genome is composed of 4 loci: 1 ) the R locus corresponding to the polymerase sequence , 2 ) the Morphogen ( M ) locus , 3 ) the C locus which corresponds to the sequences for the cis-regulatory region of the transcription factor and 4 ) the α locus , which is the morphogen gradient steepness α . Hybrids between the two lineages are constructed by independent reassortment of these loci assuming complete linkage within each locus and no linkage between them . We define a hybrid genotype by a 4 letter string where each letter corresponds to one of the loci defined above and takes one of two cases correspond to whether the allele is from the 1st line or 2nd line; for example , the hybrid rMCa corresponds to R locus having an allele from the 1st lineage , M locus with the allele from the 2nd lineage , the transcription factor ( cis-regulatory ) C locus from the 2nd lineage and α locus the allele from the 1st lineage . Note that the underlying sequence of each hybrid changes as different substitutions are accepted in each lineage; the notation only refers to alleles fixed at any point in time . We can represent all combinations of the four loci drawn from the two parents ( RMCA , RMCa , RMcA , etc . ) as points on a four-dimensional Boolean hypercube . In total there are 24 − 2 = 14 hybrids . In Fig 3 , we plot a typical time series of how the fitness of two different hybrids ( Rmca ( a & b ) and RMcA ( c & d ) changes over a divergence time μt separating a pair of lineages , for 2NκF = 1 ( a & c ) and 2NκF = 10 ( b & d ) , where μ = ℓGμ0 is the mutation rate for all base pairs in all loci . 2NκF > 1 indicates strong selection , whilst 2NκF ≤ 1 indicates weak selection where genetic drift dominates ( For reference , in human populations it has been estimated that ≈ 20 − 30% of mutations are weakly selected [29 , 30] , compared to in Drosophila < 10% [30] . ) . We see that the fitness of hybrids generally decreases in a stochastic fashion; when the log-fitness of a hybrid drops below the threshold F* ( indicated by the dashed line ) , a DMI arises as is indicated by a vertical log-fitness line ( F = −∞ ) for that hybrid . As can be seen in Fig 3 , at any given time a changing subset of the fourteen possible hybrids might be incompatible .
One important and well explored example is the evolution of transcription factor DNA binding [15 , 16 , 22 , 26 , 27 , 49 , 50] , where the genotype-phenotype map from sequence to binding affinity can be explicitly enumerated under simplifying assumptions [51 , 52] . These investigations show that for small populations dominated by genetic drift , evolution does not optimise fitness . Rather , there is a trade off between the high fitness of a small number of sequences that bind well and the exponentially larger number of sequences that bind less well . The result is the maximisation of a combination of fitness and the number of sequences that correspond to that phenotype . We can take advantage of analogies with statistical mechanics and represent this combination as the “free fitness” , where the log of the number of sequences is the “sequence entropy” of a phenotype [15 , 26 , 27 , 53] . In this formulation , the effective population size is analogous to an inverse temperature for a physical system connected to a heat-bath , where decreasing population size increases the effect of drift and the importance of sequence entropy relative to fitness . When the free fitness framework is applied to the role of transcription factor DNA binding in allopatric speciation , our previous work gave rise to a simple prediction: incompatibilities arise more quickly for smaller , drift-dominated , populations [26 , 27] , supporting previous computational studies by Tulchinsky et al . [54] , that showed decreased hybrid fitness for smaller populations . This can be understood as a result of the greater importance of sequence entropy for small populations , resulting in common ancestors with a higher drift load , which are therefore closer to incompatible regions . As a result , fewer substitutions are required for the development of hybrid incompatibilities [26 , 27] . Conversely , those transcription factor binding site pairs under weaker selection , at a fixed population size , will give rise to incompatibilities more quickly , as they are more susceptible to drift and in the common ancestor will have a larger drift load . In this paper , we examine speciation in a more realistic genotype phenotype map . For the first time we examine how incompatibilities arise in allopatry for a simple evolutionary model of developmental system drift , where a higher level organismal spatial patterning phenotype is maintained by stabilising selection , whilst the underlying molecular binding energy phenotypes and ultimately the sequences that determine them , the genotype , are allowed to drift in the evolutionary simulations . Earlier analyses of this model demonstrated the evolution of a number of non-trivial features such as a balance between fitness and sequence entropy deciding the course of evolution at small population sizes and a roughness to the fitness landscape for phenotypes which have high fitness [17] . Importantly here , unlike in previous works [26 , 27] we do not directly select for high binding affinity , but only on the organismal level phenotype , but as we discuss , we find the same population size dependence , as well as a number of other novel phenomenon to the speciation process , which would not be obtainable by modelling selection only at the level of phenotypes or genotypes . The results show that biophysics and effective population size provide a much stronger constraint than previous simple modelling of the dynamics of hybrid incompatibilities would suggest [5 , 31] . A key result we find is that small populations are characterised by a power law growth of incompatibilities with time , vs large populations a diffusive law ( discussed below ) . Thus we suggest that empirical evidence of power law growth in incompatibilities is a signature of allopatric speciation at small population sizes . The Orr model of the growth of hybrid incompatibilities predicts that incompatibilities grow as a power law of the divergence time between allopatric lineages [5 , 31] , where the exponent represents the number of genes participating in the interaction ( e . g . 2 for a 2-way incompatibility ) . The results of our model also yield this prediction , but only when populations sizes are sufficiently small . There is , however , an alternative model for the power law behaviour to the combinatoric argument made by Orr . As argued in [27] , at small population sizes , where genetic drift is dominant and there is a large drift load , common ancestor populations are poised close to the incompatibility boundary and the growth of DMIs at short times is determined by the likelihood that a few critical substitutions arrive quickly , which is given by a Poisson process; if the critical number of substitutions is K* then for short times we would expect P I ( t ) ∼ ( μ t ) K * and so given that at least n substitutions are needed for a n-way incompatibility , we would expect K* ≥ n . In this paper , we introduced a new method to decompose DMIs into their fundamental pair-wise , 3- and 4-way incompatibilities , and find that for more complex incompatibilities ( more loci involved ) the larger the exponent of their power law growth . However , we find the exponents we measure for 3- and 4-way incompatibilities are smaller than the predicted exponents of 3 and 4 respectively . We suggest this could be due , as shown in the S1 Text , to the greater number of higher order DMIs arising just by chance , leading to an overestimation of 3- and 4-way DMIs at short times , where at short times a smaller exponent corresponds to a larger number ( i . e . τn−1 > τn for τ < 1 , where τ is some dimensionless timescale ) . Examining the growth of incompatibilities at large population sizes , we see there is a characteristic negative curvature on a log-log plot , predicted theoretically by [26] , indicating that , as the number of substitutions needed for incompatibilities is large , the changes in the hybrid traits can be approximated by a diffusion process . However , we find that a simple model of diffusion does not fit the simulation data well; instead a model of sub-diffusion , that would arise if there are a number of kinetic traps giving a broad distribution of substitution times , does fit the data well . This is consistent with the finding that the genotype-phenotype map has a rough fitness landscape , which is only revealed at sufficiently large population sizes [17] . However , it is not clear whether we would observe quenched disorder and sub-diffusive behaviour with more realistic biophysical models that include a larger alphabet size with 20 amino acids and 4 nucleotides . We also find that incompatibilities arise more rapidly in smaller populations , which is an emergent effect due to the genotype-phenotype map , giving a bias in degeneracy of different phenotypes; lower fitness and less sharp patterning organismal phenotypes have many more sequences than higher fitness , sharper patterning , phenotypes . In smaller drift-dominated populations , this means there is bias towards phenotypes of small sequence entropy ( log degeneracy ) that counteracts the tendency of natural selection to favour phenotypes of high fitness . Consequently , the common ancestor in small populations is more likely to be slightly maladapted and less substitutions are needed before hybrids develop incompatibilities . These predictions are consistent with empirical evidence for an inverse correlation of speciation rates with effective population size; the net rates of diversification from phylogenetic trees [55–57] indicate smaller populations speciate more quickly , as well from inferred times for post-zygotic isolation to arise [58–60] , where for example mammals and cichlids , which have effective population sizes of order 104 [61– 64] , develop reproductive isolation more quickly than birds , which have effective populations sizes of order 106 [65] . This model and the similar results obtained for transcription factor DNA binding [26 , 27] provide a robust explanation of how stabilising selection can give rise to this population size effect in speciation , which do not require passing through fitness valleys as do models based on the founder effect [66–69] . However , the results for this genotype-phenotype map for developmental system drift are particularly noteworthy compared to the previous results on transcription factor DNA binding , as they are obtained without directly selecting for high affinity , low sequence entropy , binding phenotypes; here we only select on the organismal spatial patterning phenotype , but nonetheless we find small populations develop hybrid incompatibilities more quickly through a similar mechanism of the interplay between fitness , sequence entropy and populations size . Although studies with more complex genotype-phenotype maps will be required , we suggest this points towards a broad principle , where the specificity required of a phenotype to be functional and of high fitness will mean that it will be coded by fewer genotypes . For example , in simple models of protein stability , the empirical observation that all proteins tend to be marginally stable , can be explained by the fact that as the stability of a protein is increased the number of sequences that give that stability decreases rapidly [19]; assuming high fitness corresponds to maximum stability , this phenotype is highly specified , as only a few sequences will meet the requirement that all inter-chain interactions in the protein are favourable . Another property that emerges from this model not obtainable by simply modelling transcription factor DNA binding is that certain molecular phenotypes are more important than others in giving rise to incompatibilities . One particular feature of this model is that the selective constraints on the different molecular binding energy phenotypes emerge through the evolutionary process of stabilising selection on the organismal phenotype , and are not specified by the model . Most strikingly , and counterintuitively , the model predicts that molecular phenotypes that are under the weakest selective constraints ( but not strictly neutral ) dominate by giving rise to the earliest incompatibilities for intermediate and large population sizes . Remarkably , here this emerges as a consequence of stabilising selection on the organismal phenotype and not due to selection imposed for good binding affinity as in previous works [27] . We note that these results have been obtained by changing the population size whilst keeping the strength of selection on the organismal trait κF fixed . It would be tempting to use these results to suggest that overall those traits in a genome under weakest selection would give rise to the earliest incompatibilities and hence dominate allopatric speciation . However , the question of the how the dynamics of hybrid incompatibilities changes as the strength of selection changes is a subtle one , which we leave to future work; in this model a reduction in κF has the effect of changing the phenotypic regions of incompatibility , with non-trivial consequences . It should also be noted , the role of sequence entropy in giving a strong population size dependence to the rate of reproductive isolation; if we consider only a peaked phenotypic landscape without a sequence entropy bias , a reduction in population size would only lead to a broadening of the phenotypic distribution and not a change in the mean of the distribution , and so a much weaker effect , as the common ancestor will still be typically far from incompatible regions . On the other hand with strong ( exponential ) degeneracy biases , the mean phenotype of the common ancestor changes strongly giving the large population size effect seen , which is as demonstrated in Fig 2 . Another finding of significance is that pair-wise or 2-way DMIs dominate compared with higher order DMIs ( 3- and 4-way in this model with 4 loci ) . This is in contrast to Orr’s theoretical argument that predicts a very specific ratio of 2-way: 3-way: 4-way DMIs , equal to 12: 24: 14 , which assumes that the fraction of viable paths from the common ancestor to the current day species increases as we consider higher order DMIs [5] . This argument partly rests on the assumption that the number of inviable genotypes remains fixed as a larger number of loci are considered , which would seem a very strong assumption . Despite its simplicity , the genotype-phenotype map in this paper has many of the key features required for higher levels of epistasis , with protein-DNA binding , protein-protein binding and control of the morphogen steepness , all interacting in a non-linear fashion to produce a single gene expression patterning phenotype and so there is clearly the potential for the Orr prediction to be verified; in contrast , we find the converse and our results show there is no bias towards 3-way DMIs , in fact showing instead that the ratio of 2-way to 3-way DMIs is at short times many orders of magnitude larger . This suggests , in this simple , but still relatively complex model , that biophysical constraints are far more important than a purely combinatorial argument would suggest . Evidence could be obtained from more detailed studies similar to [6 , 70] , where a power law with an exponent greater than 2 would indicate higher-order DMIs are dominant; currently this evidence suggests a quadratic growth law , however , a study with more time-points or species-pairs would provide more confidence . An alternative approach would be to look for linkage disequilibrium between unlinked regions of hybrid genomes , such as was found with hybrids of two species of swordtail fish [71] , and though computationally challenging and requiring large numbers of parallel datasets , compare this against evidence for pervasiveness of higher order epistasis . Although recent results of [72] , would seem to contradict our conclusions , their finding of extensive complex epistasis relates to higher order interactions between sites within a single locus , coding for protein stability or enzymatic activity , whereas our work relates to epistasis between multiple loci . Similarly , the results of hybrid incompatibilities within RNA molecules [73] , which show a ‘spiralling complexity’ of DMIs would appear to be of limited biological relevance to allopatric speciation in higher organisms , as these are related to epistasis within a single locus , which are unlikely to segregate into different recombinants in a hybrid . Finally , for small populations we find clustering in the behaviours of growth of different types of DMIs , in particular , 3-way DMIs ( S1 Text ) , which can be explained by the different sequence entropy constraints on different molecular phenotypes . This degeneracy is then lifted at larger population sizes and each n-way DMI takes on a different identity in their pattern of growth; this has strong analogy to physical phenomenon in statistical physics where constraints of symmetry dominate at large temperatures , in a regime where noise is important , but at smaller temperatures this symmetry is broken . A clear future direction to investigate would be the effect of multiple transcription factors binding to enhancer regions to control gene expression [74–76] in large gene networks , where there is potential scope for complex epistasis across many loci coding for a large number of transcription factors . However , as our results show , despite the possibility and a prior expectation of a larger number of triplet interactions , pair-wise interactions dominate; for complex transcriptional control , if pair-wise interactions between proteins , and proteins and DNA dominate , for example in determining the binding affinity of transcriptional complexes , then our conclusions would hold . As we broaden the scope to large gene regulatory networks , there is no strong and direct empirical evidence for pervasive higher order epistasis in their function , which could give rise to higher order incompatibilities being dominant [77] . Specifically , although there is evidence that higher order incompatibilities have arisen in natural populations [78–81] , nonetheless a survey of these findings suggest there is no evidence for their dominance [81] as would be predicted by Orr and would be consistent with our findings that point towards biophysics providing a stronger constraint . Overall , our results point to a basic principle , where developmental system drift or cryptic variation [7 , 10 , 11 , 43] , play a key role in speciation; basic body plans or phenotypes are conserved , but co-evolution of the components and loci of complicated gene regulatory networks can change differently in different lineages , giving incompatibilities that grow in allopatry . Here , we suggest a universal mechanism , where the rate of growth of incompatibilities is controlled by the drift load , or distribution of phenotypic values , of the common ancestor , which in turn is determined by a balance between selection pushing populations towards phenotypes of higher fitness and genetic drift pushing them towards phenotypes that are more numerous ( higher sequence entropy ) ; this basic mechanism would predict in general that traits under weaker selection will dominant the initial development of reproductive isolation . In particular , although in principle more complicated regulation could give rise to more complex patterns of epistasis [5] , our findings suggest that more simple , pair-wise , incompatibilities dominate the development of reproductive isolation between allopatric lineages under stabilising selection .
We model the binding energies of proteins to DNA using the “two-state” approximation [51 , 52] , which assumes that the binding energy of each amino acid-nucleotide interaction at the binding interface is additive and to a good approximation controlled by the number of mismatches , which each have the same penalty in binding affinity . The various protein-DNA binding energies in the main text are then given by the Hamming distance between the respective sequences . We assume these energies E are measured relative the background free energy of specific and non-specific binding all other sites in the genome , such that the probability of a given transcription factor being bound to a single site is p = 1/ ( 1+e ( E−μ ) /kBT ) , where μ is the chemical potential ( ∼log ( concentration ) ) of the TF [17 , 26 , 82] . For example , the binding energy between the morphogen ( M ) and the first binding site ( B ) is given by E M B = ϵ p d ρ ( g m , g B ) ( 3 ) where ρ ( gm , gB ) is the Hamming distance between the protein binding sequence ( gm ) for the morphogen and the sequence for a first regulatory binding site ( gB ) , where ϵpd is the cost in energy for each mismatch . We assume ϵpd = 2kBT as a typical value for the mis-match energy , which are found to be in the range 1−3kBT [51 , 52] . Similarly the co-operative protein-protein binding energy , for example between RNAP and the morphogen is E ˜ R M is E ˜ R M = ϵ p p ( ℓ p p - ρ ( g R , g M ˘ ) ) ( 4 ) where gR is the sequence involved in protein-protein interactions for the polymerase , and g M ˘ represents the equivalent binary sequence for the morphogen , flipped about its centre , which mimics the chirality of real proteins and prevents the co-operative stability from always being maximum for homo-dimers . The parameter ϵpp is the stability added for each favourable hydrophobic interaction between amino acids , which we assume to be ϵpp = −kBT . Given ℓpp = 5 this gives interactions consistent with typical literature values of −2 to −7kBT for hydrophobic interactions between proteins [28 , 83] . The morphogen concentration profile is approximately exponential; the exact profile we use is [ M ] ( x ) = [ M 0 ] cosh ( α ( x - L ) ) sinh ( α L ) ( 5 ) where this arises from solving the reaction-diffusion equation with reflecting boundary conditions and is valid for all α . To calculate this probability , we use the canonical ensemble of statistical mechanics , for which the partition function Z is most simply expressed in terms of a spin-like variable , which represents the occupation of each binding site , σ j = { 0 , R , M } , Z = ∑ σ P ∑ σ B e - ( E σ P P + E σ B B - μ σ P - μ σ B + E ˜ σ P σ B ) / k B T with E 0 j = 0 , E ˜ i i ′ = 0 , for either i = 0 or i′ = 0 and μ0 = 0 , where μ σ j = k B T ln [ σ j ] represents the chemical potentials of the protein species at the jth binding site with [σj] being the concentration of species σj and 0 represents a free binding site . Formally this construction is known as a 3−state Pott’s model . So given a ‘genome’ G = [gR , gr , gM , gm gP , gB] from which the protein-DNA and protein-protein binding energies are calculated ( Eij and E ˜ i i ′ , respectively ) , pRP is given schematically by pRP=p ( σB−R↱ ) =1Z ( ( 0−R↱ ) + ( M−R↱ ) + ( R−R↱ ) ) ( 6 ) where , for example , ( M - R ↱ ) ≡ e - ( E R P + E M B - μ R - μ M + E ˜ R M ) / k B T is the Boltzmann factor for co-operative binding of the morphogen and RNAP to R . Note that we ignore in the partition sum protein-protein binding when not bound to DNA , since these co-operative binding energies tend to be relatively weak compared to DNA binding . We use a kinetic Monte Carlo scheme to simulate the evolutionary process for the genome G and α on two independent lineages , assuming a fixed effective population size of N , and that we are in the regime of small effective population size ( ℓG μ0N ≪ 1 , where μ0 is the base-pair mutation rate ) . This means the population is represented by a single fixed sequence ( or number for α ) for all of the loci at each time-point , where effectively mutations are either instantly fixed or eliminated . Specifically , we use the Gillespie algorithm [84] , to simulate evolution as a continuous time Markov process; at each step of the simulation the rate of fixation of all one-step mutations from the currently fixed alleles ( wild type ) is calculated , and one of these mutations is selected randomly in proportion to their relative rate . Time is then progressed by K−1 ln ( u ) , where K is the sum of the rates of all one-step mutants and u is a random number drawn independently between 0 and 1 , which ensures the times at which substitutions occur is Poisson distributed , as we would be expected for a random substitution process . The rates are based upon the Kimura probability of fixation [85]: k = μ 0 N 1 - e - 2 δ F 1 - e - 2 N δ F ≈ μ 0 2 N δ F 1 - e - 2 N δ F , ( 7 ) where δF is the change of fitness of a mutation at a particular location , given by fitness function defined in the main text , and μ0N is the rate at which mutations arise in the population; the latter approximation in Eq 7 assumes δF ≪ 1 . Note that although in the simulations we use the full form for the fixation probability , fitness effects are typically small ( δF ≪ 1 ) in the simulations , so the substitution rates only depend on the population-scaled fitness changes 2NδF which , for a given mutation , is proportional to 2NκF . Finally , we allow continuous ‘mutations’ in the morphogen steepness parameter α , chosen from a Gaussian distribution of standard deviation δα = 0 . 5 and assign it an 10 effective base-pairs , which are used when assigning relative weight in the kinetic Monte Carlo scheme , where the total number of base-pairs is ℓG = 60 . We determine the Malthusian or log fitness of the spatial gene regulation , from the resulting concentration profile [TF] ( x ) by use of a functional that promotes expression of theTF in the anterior half , whilst penalising expression in the posterior half , with truncation selection below a critical value F*: F = { κ F ln ( W ) if κ F ln ( W ) > F * - ∞ if κ F ln ( W ) < F * ( 8 ) where , W [ [ T F ] ( x ) ] = ∫ 0 L / 2 [ T F ] ( x ) d x - ∫ L / 2 L [ T F ] ( x ) d x L 2 max x { [ T F ] ( x ) } . ( 9 ) where we use a value of F* = −1 . 6 × 10−3 , which corresponds a value of W ≈ 0 . 2 , when κF = 10−3 . Strictly , an inviability on a lineage or a hybrid should correspond to W = 0 or F* = −∞ , however , these values were chosen to so that a reasonable number of incompatibilities arise in a simulation; for comparison the typical maximum of the integral W ≈ 0 . 6 . In this paper , we explore how the changing the population scaled strength of selection ( 2NκF ) affects the rate of reproductive isolation , by keeping κF fixed and varyingN accordingly . Note that although here the exact form of the fitness is slightly different to the one used in [17] , the qualitative behaviour is the same ( S1 Text ) . The speciation simulations consist of two replicate simulations starting with the same common ancestor and with the same fitness function . We draw the common ancestor from the equilibrium distribution for G and α . To do this we start from a random initial genome , and run one long simulation for 100 , 000 substitutions for a fixed scaled population size 2NκF , in order to effectively equilibrate the system ( typically 10 , 000 substitutions are required to adapt to an ensemble of fit states ) . This represents a reference equilibrium state; different random draws from the equilibrium distribution then consist of running the simulation for a further 100 substitutions . Given a pattern of hybrid incompatibilities , for example , as shown in Fig 4 , if there is a 2-way DMI ( e . g . between C and α loci , which we denote ICa ) , then all four hybrid-genotypes containing this DMI ( e . g . RMCa , RmCa , rMCa , rmCa ) are inviable; these genotypes define a two-dimensional subspace ( or face ) of the hypercube . Similarly , the points ( e . g . rmcA , RmcA , which we denote ImcA ) containing a 3-way DMI form a one-dimensional subspace ( or line ) , while a 4-way DMI takes up only a single point in the four-dimensional hypercube . These different 2-way , 3-way and 4-way DMIs are the fundamental incompatibility types which we seek to explain the pattern of hybrid inviable genotypes observed , for example , as in Fig 4a . However , this decomposition is hugely underdetermined , as there are only 24 − 2 = 14 possible hybrids ( not including the well-adapted genotypes of line 1 and line 2 ) and a total of Imax = 3L + 1 − 2L+1 = 50 , different fundamental incompatibilities , for L = 4 loci . This arises as the total number of n−point DMIs is ( 2 n - 2 ) ( L n ) , as there are ( L n ) combinations ofN loci amongst L total loci and then considering a binary choice of alleles across both lines , there are a total of 2n allelic combinations or states , 2 of which are the fit allelic combinations where all alleles come from one lineage or the other giving 2n − 2 . For example , between each pair of loci there are 22 − 2 = 2 mismatching combinations of alleles ( e . g . rM and Rm ) that could give DMIs and ( L 2 ) = L ( L - 1 ) / 2 = 6 pairwise interactions . A similar argument would give a total of 24 3-way DMIs as there are 23 − 2 = 6 mismatching combinations of alleles at 3 loci ( e . g . , excluding rmc and RMC ) and ( L 3 ) = 4 3-way interactions and similarly , 14 ( L 4 ) = 14 for 4-way interactions . In total , the max number of DMIs is I m a x = ∑ n = 2 L ( 2 n - 2 ) ( L n ) = 3 L + 1 - 2 L + 1 , which for L = 4 loci is Imax = 50 . The approach we take is to find only those combinations of fundamental DMIs that have the smallest total number that can explain the pattern of hybrid incompatibilities , which from a Bayesian perspective would have the smallest Occam factors [86]; for instance , as shown in Fig 4b the list of 6 incompatible hybrid genotypes rmCa , rMCa , RmCa , RMCa , Rmca , RMca , shown by red crosses , can be explained most parsimoniously by three different minimal combinations of fundamental DMIs , each with only 2 DMIs ( see main text ) .
|
The process of speciation is of fundamental importance to the field of evolution as it is intimately connected to understanding the immense bio-diversity of life . There is still relatively little understanding of the underlying genetic mechanisms that give rise to hybrid incompatibilities with results suggesting that divergence in transcription factor DNA binding and gene expression play an important role . A key finding from the field of evo-devo is that organismal phenotypes show developmental system drift , where species maintain the same phenotype , but diverge in developmental pathways; this is an important potential source of hybrid incompatibilities . Here , we explore a theoretical framework to understand how incompatibilities arise due to developmental system drift , using a tractable biophysically inspired genotype-phenotype for spatial gene expression . Modelling the evolution of phenotypes in this way has the key advantage that it mirrors how selection works in nature , i . e . that selection acts on phenotypes , but variation ( mutation ) arise at the level of genotypes . This results , as we demonstrate , in a number of non-trivial and testable predictions concerning speciation due to developmental system drift , which would not be obtainable by modelling evolution of genotypes or phenotypes alone .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"gene",
"regulation",
"regulatory",
"proteins",
"dna-binding",
"proteins",
"dna",
"transcription",
"developmental",
"biology",
"speciation",
"transcription",
"factors",
"molecular",
"development",
"population",
"biology",
"thermodynamics",
"entropy",
"proteins",
"gene",
"expression",
"physics",
"morphogens",
"population",
"metrics",
"biochemistry",
"population",
"size",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"evolutionary",
"biology",
"evolutionary",
"processes"
] |
2019
|
Biophysics and population size constrains speciation in an evolutionary model of developmental system drift
|
Late endosome-resident interferon-induced transmembrane protein 3 ( IFITM3 ) inhibits fusion of diverse viruses , including Influenza A virus ( IAV ) , by a poorly understood mechanism . Despite the broad antiviral activity of IFITM3 , viruses like Lassa virus ( LASV ) , are fully resistant to its inhibitory effects . It is currently unclear whether resistance arises from a highly efficient fusion machinery that is capable of overcoming IFITM3 restriction or the ability to enter from cellular sites devoid of this factor . Here , we constructed and validated a functional IFITM3 tagged with EGFP or other fluorescent proteins . This breakthrough allowed live cell imaging of virus co-trafficking and fusion with endosomal compartments in cells expressing fluorescent IFITM3 . Three-color single virus and endosome tracking revealed that sensitive ( IAV ) , but not resistant ( LASV ) , viruses become trapped within IFITM3-positive endosomes where they underwent hemifusion but failed to release their content into the cytoplasm . IAV fusion with IFITM3-containing compartments could be rescued by amphotericin B treatment , which has been previously shown to antagonize the antiviral activity of this protein . By comparison , virtually all LASV particles trafficked and fused with endosomes lacking detectable levels of fluorescent IFITM3 , implying that this virus escapes restriction by utilizing endocytic pathways that are distinct from the IAV entry pathways . The importance of virus uptake and transport pathways is further reinforced by the observation that LASV glycoprotein-mediated cell-cell fusion is inhibited by IFITM3 and other members of the IFITM family expressed in target cells . Together , our results strongly support a model according to which IFITM3 accumulation at the sites of virus fusion is a prerequisite for its antiviral activity and that this protein traps viral fusion at a hemifusion stage by preventing the formation of fusion pores . We conclude that the ability to utilize alternative endocytic pathways for entry confers IFITM3-resistance to otherwise sensitive viruses .
Fusion of enveloped viruses with the host cell membrane is a key step leading to infection . Viral fusion is initiated upon interactions between virus surface glycoproteins and cellular receptor ( s ) and/or upon the reduction in pH that follows endocytosis ( reviewed in [1–3] ) . The extensive conformational changes that ensue in the viral glycoproteins promote fusion between viral and cellular membranes [4–6] . There is strong evidence that viral fusion—and membrane fusion in general—proceeds through a hemifusion intermediate defined as a merger of two contacting membrane leaflets without additional merger of distal leaflets that results in the formation of a fusion pore [6–8] . Accordingly , hemifusion is manifested as lipid mixing between viral and host membranes without viral content release , while full membrane fusion entails mixing of distinct aqueous contents delimited by the two membranes [4–6] . It has been demonstrated that sub-optimal conditions for membrane fusion including low density of viral glycoproteins , reduced temperature , and—where applicable—insufficiently acidic pH , favor dead-end hemifusion that does not progress to full fusion [9–12] . Thus , the progression to full viral fusion that culminates in the release of nucleocapsid into the cytoplasm is largely dependent on local conditions . The viral envelope glycoproteins responsible for mediating membrane fusion are targets for neutralizing antibodies and virus entry inhibitors . In addition , new innate restriction factors inhibiting virus fusion have been discovered in recent years [13–15] . Among these factors is the family of small interferon-induced transmembrane proteins ( IFITMs ) that exhibits broad-range of antiviral activity [15–17] . This family includes IFITM1 , which localizes predominantly at the plasma membrane , as well as IFITM2 and IFITM3 , which contain an endocytic signal in their cytoplasmic N-terminal domain and thus localize to late endosomal and lysosomal membranes [18–21] . IFITMs effectively block entry of many unrelated enveloped viruses , including orthomyxoviruses ( influenza A virus , IAV ) , paramyxoviruses ( Respiratory Syncytial Virus , RSV ) , flaviviruses ( Dengue , West Nile ) , filoviruses ( Marburg , Ebola ) , and coronaviruses ( SARS ) [15–17 , 20 , 22–25] . IFITM3 alone is responsible for the bulk of antiviral effects of interferon in cell culture [15] . Importantly , mice lacking the ifitm3 gene more readily succumb to IAV and RSV infection than control mice [26 , 27] . There are , however , viruses that are resistant to IFITM-mediated restriction . Murine Leukemia Virus ( MLV ) , Old and New World arenaviruses ( Lassa Virus and Junin Virus , respectively ) , as well as several enveloped DNA viruses , are not affected by IFITMs [15 , 28 , 29] . The mechanism by which IFITMs inhibit fusion of most viruses , while sparing others , is not understood . We and others have shown that IFITM expression does not elevate the overall endosomal pH [15–19 , 22 , 30 , 31] and , thus , should not block acid-triggered refolding of viral fusion proteins that initiate membrane fusion . Clues regarding the antiviral mechanisms of IFITMs come from their subcellular distribution which tend to correlate with IFITM’s potency against different viruses . IFITM2 and -3 better restrict viruses entering from late endosomes , while IFITM1 tends to be more effective against viruses that are thought to fuse with the plasma membrane or with early endosomes ( reviewed in [17] ) . Indeed , expression of an IFITM3 mutant that redistributes the late endosome/lysosome-resident protein to the cell surface abolishes antiviral activity against IAV [32] . There are , however , exceptions to this rule . The fact that IFITM1 outperforms IFITM3 in restricting EBOV fusion [25] highlights the importance of cellular trafficking , as opposed to the steady state distribution , for antiviral activity . Also , a relatively weak IAV restriction exhibited by an IFITM1 chimera containing the N-terminal domain of IFITM3 that localizes to late endosomes suggests a role for other factors in addition to appropriate subcellular localization [21] . The most popular view of the mechanism of IFITM’s antiviral activity is that these proteins create “tough membranes” that are not conducive to fusion [17 , 18 , 22] . Two principal models for membrane stiffening by IFITMs have been proposed–a direct effect on the membrane in the immediate proximity of these proteins [19 , 25 , 33–35] that could involve changing the membrane fluidity and/or curvature [22 , 33 , 35] , and an indirect effect through altering the lipid composition of endosomes [18] . Several lines of evidence support the proximity-based antiviral activity of IFITMs . First , as discussed above , there is a general correlation between the subcellular localization of IFITMs and their potency against viruses entering from distinct cellular compartments ( reviewed in [17] ) . Second , IFITM3-mediated restriction , but not restriction by the plasma membrane-resident IFITM1 , can be bypassed by forcing virus fusion with the plasma membrane [25 , 30] . Third , IFITM incorporation into the viral membrane effectively inhibits fusion/infectivity [34 , 36–38] . On the other hand , IFITM3 has been reported to bind to and inhibit the function of vesicle-associated membrane protein-associated protein A ( VAPA ) [18] , the master regulator of endosome-ER lipid transport . While this model has been disputed by several groups [30 , 35] , a recent study provided evidence for the antiviral effect of cholesterol accumulation in late endosomes/lysosomes and confirmed accumulation of cholesterol in these compartments upon IFITM3 expression [39] . It thus remains unclear whether IFITMs must be present at the sites of virus fusion to block virus entry or affect fusion indirectly , by dysregulating lipid transport or metabolism . We have previously shown that IFITM3 does not restrict the lipid-mixing ( hemifusion ) stage of viral fusion , but rather inhibits the formation of a fusion pore [30] . However , the inability to directly visualize IFITM3 in the context of virus entry into live cells precluded us from assessing whether this factor blocks fusion through a proximity-based mechanism . Here , we overcame this limitation by constructing a functional fluorescent IFITM3 protein and imaging virus co-trafficking and fusion with endosomal compartments in cells expressing this protein . Comparison of entry and fusion of IFITM3-sensitive ( IAV ) and–resistant ( LASV ) viruses by 3-color live cell imaging revealed that IAV enters into and remains trapped within endosomes enriched in fluorescent IFITM3 , where viruses underwent hemifusion but failed to complete the fusion reaction . In contrast , LASV particles entered and fused with endosomes devoid of IFITM3 , implying that LASV escapes restriction by utilizing an endocytic pathway distinct from that employed by IAV . Collectively , our results provide strong support to a proximity model by which presence of IFITM3 at the preferred sites of virus entry restricts viral fusion .
To assess the co-distribution of viruses with IFITM3 at the time of fusion in live cells , we generated fluorescently-tagged IFITM3 protein . Linear N- or C-terminal fusions of EGFP ( or similar fluorophores ) with IFITMs render IFITMs nonfunctional . However , we found that the coding sequence of EGFP inserted into the N-terminal region of IFITM3 , predicted to reside in the cytoplasm [40–42] , generates a functional protein ( Fig 1 ) . First , to verify correct subcellular distribution of the fluorescent construct , we co-expressed IFITM3-iEGFP ( iEGFP stands for “internal” EGFP ) with an N-terminally myc-tagged IFITM3 in HeLa cells and confirmed their colocalization by fixing and immunostaining the cells ( Fig 1B ) . Extensive colocalization between IFITM3-iEGFP and myc-IFITM3 suggests that subcellular localization of IFITM3 is not perturbed by incorporation of an EGFP tag . To test the functionality of the fluorescent construct , control A549 cells transduced with an empty vector ( Vector ) and cells transduced with IFITM3-iEGFP were infected with varied doses of influenza A/WSN/33 virus and the resulting infection measured by immunostaining cells for HA antigen . A549 cells were selected because they express very low endogenous levels of IFITM3 [15 , 30 , 31] . Cells expressing the EGFP-labeled IFITM3 were consistently much more resistant to influenza infection than control cells ( Fig 1C ) . Microscopic analysis revealed that cells expressing intermediate to high levels of IFITM3-iEGFP did not stain for HA antigen ( Fig 1D ) , demonstrating that IFITM3-iEGFP protects cells from IAV infection , similar to unlabeled IFITM3 [15 , 30] . To facilitate multi-color live cell imaging of IFITM3 with fluorescently labeled viruses , we replaced the internal EGFP tag with cyan mTFP1 or bright green mNeonGreen protein . We next examined the ability of fluorescent IFITM3 constructs to inhibit IAV fusion using HIV-1 particles pseudotyped with influenza HA and NA proteins from the H1N1 A/WSN/33 strain ( designated as IAVpp ) and carrying the β-lactamase-Vpr ( BlaM-Vpr ) chimera , as described in [30] . A549 cells transduced with an empty vector , unlabeled IFITM3 , IFITM3-imNG , or IFITM3-imTFP1 were inoculated with IAVpp , and the extent of viral fusion was measured after 2 h at 37°C based on the resulting cytosolic BlaM activity [43 , 44] . Compared to the Vector control , IFITM3 severely restricts IAVpp fusion , as shown previously [30] . IFITM3-imNG and IFITM3-imTFP1 proteins were expressed in A549 cells at levels comparable to that of untagged IFITM3 , as determined by Western blot analysis of respective cell lysates ( Fig 1F ) , and also potently inhibited IAVpp fusion ( Fig 1E ) . Comparable expression of the two fluorescent constructs is further supported by live cell fluorescence microscopy analysis ( S1 Fig ) . Taken together , our results demonstrate both the appropriate subcellular localization and antiviral activity of fluorescently labeled IFITM3 constructs . Lipid mixing between two membranes is a necessary but not sufficient condition for complete fusion; lipids can diffuse through a hemifusion intermediate , without opening of a fusion pore ( e . g . , [11 , 30 , 45 , 46] ) . We have previously shown that IFITM3 expression does not inhibit lipid mixing ( hemifusion ) between IAV and host endosomal membranes , but rather interferes with the formation of fusion pores [30] . The ability to visualize the distribution of functional fluorescent IFITM3 in living cells enables spatiotemporal analysis of viral fusion restriction . To test lipid mixing activity in the context of endosomes containing fluorescent IFITM3 , we co-labeled infectious IAV ( A/PR/8/34 H1N1 ) with a self-quenching concentration of the lipophilic dye SP-DiI18 and with the amine-reactive far-red Alexa Fluor647 dye ( AF647 ) that labels surface glycoproteins of the virus [30] . As shown in the schematic diagram , single-virus hemifusion with an endosome can be detected by the appearance of bright SP-DiI18 spots resulting from dilution of this dye within an endosomal membrane ( Fig 2A ) . Double-labeled IAV were incubated with cells at 4°C for 20 min , followed by the addition of pre-warmed Live Cell Imaging Buffer ( LCIB ) , and imaging continued at 37°C . Representative snapshots from the time-lapse movie illustrate the single IAV SP-DiI18 ( colored green ) dequenching event around 25 min post-infection of Vector cells , indicating redistribution of the dye into an endosomal membrane ( Fig 2B , S1 Movie ) . Fluorescent traces obtained by single IAV tracking show an increase in SP-DiI18 intensity over time , whereas the reference AF647 signal ( red ) remains relatively constant ( Fig 2C ) . From these traces , the time required for complete dequenching ( Δt ) and the extent of dequenching ( ratio of the initial and final mean intensities If/Ii ) can be determined ( Fig 2C ) . To evaluate whether lipid mixing between IAV and IFITM3-containing compartments occurs , AF647 and SP-DiI18 labeled viruses were incubated with A549 cells expressing IFITM3-imNG , and virus entry/fusion monitored by three-color live cell microscopy ( Fig 2D ) . As expected , fluorescent IFITM3 primarily localized to endosomes that exhibited retrograde and anterograde movement in living cells . The dynamics of IFITM3-imTFP1 intracellular transport is illustrated by S2 Movie . Snapshots of a single IAV show the hemifusion event that occurs within an IFITM3-imNG endosome ( IFITM3 ) , with the onset of SP-DiI18 dequenching detected at around 32 min ( Fig 2E , S3 Movie ) . The overall kinetics of onset of lipid mixing was not affected by IFITM3 expression ( S2A Fig ) . Fluorescence traces with levels of fluorescence intensity from IFITM3 ( blue ) , SP-Dil18 ( green ) , and AF647 ( red ) are shown with the time to SP-DiI dequenching ( Δt ) and dequenching ratio of the final and initial mean intensities ( Fig 2F ) . It is clear that IAV HA-mediated lipid mixing is not inhibited by accumulation of IFITM3 in endosomes . We next asked whether the presence of fluorescent IFITM3 affects the rate or extent of lipid exchange between IAV and the limiting membrane of a vesicle . The fold of SP-DiI18 dequenching was measured by calculating mean ratios of single IAV SP-DiI18 signals after and before dequenching ( If/Ii ) in A549 Vector cells and in A549-IFITM3-imNG cells occurring within IFITM3+ punctae ( as in Fig 2F ) . The extent of SP-DiI18 dequenching was independent of virus colocalization with IFITM3-imNG endosomes at the time of lipid mixing ( Fig 2G ) . Since the extent of dequenching of lipophilic dyes is proportional to their fold-dilution ( e . g . , [47 , 48] ) , this result indicates that the average size of recipient endosomes is the same in control and IFITM3-expressing cells . However , the SP-DiI18 dequenching time ( Δt ) was significantly longer in IFITM3+ endosomes compared to Vector cells ( Fig 2H ) , consistent with a hemifusion connection that is more restrictive for lipid diffusion in IFITM3+ endosomes compared to control cells . The slower redistribution of SP-DiI18 to endosomes enriched in IFITM3-imTFP1 is consistent with our previous conclusion that these events represent IAV hemifusion but not full fusion [30] . Indeed , lipid diffusion through both leaflets of a fusion pore is expected to be faster than diffusion through contacting leaflets of a hemifusion intermediate [49] . In another example , IAV SP-DiI18 dequenching can also occur with a bi-phasic increase in intensity , suggesting either transient fusion pore closure or representing the transition from a restrictive hemifusion structure that attenuates lipid diffusion to a fusion pore ( S2 Fig ) . To visualize single IAV fusion in A549 cells , we pseudotyped the HIV-1 core with H1N1 HA and NA glycoproteins , as previously described [30] . IAVpp was labeled with a bi-functional mCherry-2xCL-YFP-Vpr construct , with a 2xCL tandem cleavage site for the viral protease that is cleaved during virus maturation , generating a free mCherry and a core-associated YFP-Vpr [50] . Virus fusion is detected based upon the release of mCherry into the cytoplasm through a fusion pore , while the YFP-Vpr marker remains in the viral core ( Fig 3A ) . Labeled IAVpp were bound to A549 Vector cells in the cold by spinoculation ( see Methods ) . Virus fusion was synchronously initiated by adding pre-warmed LCIB and visualized by two-color live cell microscopy for 2 hours with mCherry and YFP signals acquired every 6 seconds . Single IAVpp entered and fused with A549 cells , in agreement with our published data [30] . Time series images show the initial trafficking of a representative mCherry/YFP-Vpr labeled virus until fusion occurs , as evidenced by the loss of mCherry fluorescence , while the YFP-Vpr signal remains relatively constant ( Fig 3B and 3C , S4 Movie ) . Considering that the overwhelming majority of the IAV lipid mixing events occur in IFITM3-containing endosomes ( Fig 2 ) and based upon our previous observation that IFITM3 expression inhibits single IAVpp fusion [30] , we hypothesized that a sufficiently high local IFITM3 concentration is required for restriction of IAV fusion . To test this hypothesis , we synchronized IAVpp entry into A549-IFITM3-imTFP1 cells , as described above , and monitored mCherry/YFP-Vpr-labeled viral particles using three-color live cell imaging . Time series images show the IAVpp entering the IFITM3+ compartment and co-trafficking without undergoing fusion ( Fig 3D ) . The fluorescence traces obtained by single virus tracking confirm entry ( indicated by the red arrow ) into an IFITM3+ endosome . At a later time , the virus-carrying endosome encounters and merges with another IFITM3+ compartment , as indicated by the second red arrow around 36 min ( Fig 3E , S5 Movie ) . This pseudovirus did not fuse ( release mCherry ) for as long as the time-lapse imaging was performed . Co-trafficking of IAVpp with an IFITM3+ compartment can be visualized by examining 3D trajectories of the virus and relevant endosomes , which shows IAVpp co-trafficking with the first and then the second endosomal compartments containing IFITM3-imTFP1 ( Fig 3F ) . Importantly , analysis of 6309 particles did not reveal a single viral fusion event occurring after extensive IAVpp co-trafficking with an IFITM3+ compartment . In addition to tracking IAV particles that co-traffic with IFITM3+ compartments for an extended period of time , transient encounters with IFITM3+ compartments that did not inhibit subsequent IAVpp fusion were also observed . An example trace of an IAV pseudovirus shows a brief ( ~40 sec ) apparent co-localization with an IFITM3+ compartment at around 23 minutes ( Fig 3G and 3H , S6 Movie ) . Examination of 3D trajectories demonstrate that the particle does not significantly colocalize/co-traffic with the IFITM3+ endosome and that fusion occurs at around 31 minutes with no above-background IFITM3-imTFP1 signal ( Fig 3H and 3I ) . Another example of false colocalization of IAVpp with an IFITM3+ endosome is shown in S3 Fig and S7 Movie . The viral particle appears to transiently colocalize with an IFITM3+ endosome when visualized in 2D , but particle tracking performed in 3D shows that the viral particle and the IFITM3+ endosome traffic in different Z-planes . These observations suggest that a transient and chance encounter of a virus carrying endosome with an IFITM3+ endosome is not sufficient to restrict fusion . In contrast , fusion is blocked after sustained and prolonged IAVpp co-trafficking with an IFITM3+ endosome , which strongly implies that the virus is being carried by an IFITM3-enriched compartment . Further analysis of time-lapse acquisitions performed in at least 15 independent experiments shows that , on average , 2 . 2% of IAVpp fuse in Vector cells and none of the 6309 analyzed particles in IFITM3+ endosomes underwent fusion ( Fig 3J ) . Our results thus demonstrate , for the first time , that the presence of IFITM3 in the endosomes carrying the virus is key to restriction of IAV fusion . The unimpeded lipid mixing between IAV and IFITM3+ endosomes , together with the lack of viral content release , strongly imply that IFITM3 traps the IAV fusion at a hemifusion stage by blocking the formation of a small fusion pore ( in agreement with our previous study [30] ) . Inhibition of IAVpp fusion after sustained co-trafficking with IFITM3-imTFP1-enriched endosomes may occur through a direct block of viral fusion by the restriction factor . Alternatively , IFITM3 may indirectly interfere with IAV fusion by altering the properties of endosomes , such as the luminal pH . The overall acidity of IFITM3+ endosomes was assessed by loading A549-IFITM3-imTFP1 cells with the acidic compartment marker , LysoTracker Red ( S4 Fig ) . Imaging of fixed cells shows strong cytoplasmic colocalization between IFITM3+ endosomes and LysoTracker Red positive compartments , with only a small fraction of peripheral IFITM3+ endosomes lacking a detectable LysoTracker Red signal ( S4A Fig , Inset ) . Analysis of multiple fields of view confirms that most IFITM3+ endosomes , with the exception of a small number of peripheral endosomes , accumulate the lysosomal marker ( S4B Fig ) . This finding is consistent with progressive acidification of early IFITM3-containing endosomes through a maturation process and thus supports the notion that virus-carrying late IFITM3+ endosomes are acidic . To further test whether IFITM3+ compartments are otherwise permissive for viral fusion , we sought to render IFITM3-imTFP1 inactive by pretreating cells with Amphotericin B ( AmphoB ) , which is known to antagonize the antiviral activity of IFITM3 ( [35] and Fig 4A ) . We also used the inactive oligomerization-defective IFITM3 mutant , with alanine substitutions at F75 and F78 ( denoted 2M ) [21] . As expected , the 2M-IFITM3-imTFP1 mutant did not inhibit IAVpp fusion ( Fig 1E ) , in spite of being expressed at a level comparable to IFITM3 and IFITM3-imTFP1 ( Fig 1F ) . We next probed the ability of single IAVpp to fuse with IFITM3+ compartments under conditions that rescue bulk IAVpp fusion . A549-IFITM3-imTFP1 cells were infected with IAVpp labeled with mCherry-2xCL-YFP-Vpr , as above , in the presence of 1 μM AmphoB . As seen with a bulk fusion assay ( Fig 4A ) , single IAVpp fuses with IFITM3+ endosomes in the presence of AmphoB . Representative single virus images show the entry and subsequent co-trafficking of an IAV particle with an IFITM3+ compartment and fusion within the compartment , as indicated by the sudden loss of mCherry ( Fig 4B and 4C , S8 Movie ) . Of the 23 total fusion events that occur in IFITM3-imTFP1 cells treated with AmphoB , 7 particles co-traffic and fuse with IFITM3+ compartments ( Fig 4G ) . In contrast to IFITM3-imTFP1 expressing cells in the presence of AmphoB , IAVpp exclusively fused at sites devoid of the mutant IFITM3 in 2M-IFITM3-imTFP1 cells ( Fig 4D and 4E , S9 Movie ) . None of the IAVpp fusion events of the total 4442 particles annotated in 2M-IFITM3 cells co-traffic with IFITM3+ compartments . Fig 4D and 4E illustrates this phenomenon , whereby an IAVpp particle fuses within a 2M-IFITM3-imTFP1 cell but does not co-traffic with appreciable local 2M-imTFP1 maxima . These results suggest that loss of antiviral activity of the F75/78A IFITM3 mutant may be due to its altered subcellular distribution that prevents co-trafficking with IAV . This is in contrast to AmphoB , which renders wild-type IFITM3-imTFP1 inactive without affecting its trafficking pathways . Of note , both conditions that rescued the IAVpp fusion delayed the fusion kinetics relative to untreated cells expressing IFITM3-imTFP1 ( S5 Fig ) , indicating a global effect on the rate of virus endocytosis and entry into permissive compartments . Together , the above results support the notion that IFITM3 inhibits IAV fusion through a proximity-based mechanism–by co-trafficking with the virus and accumulating in compartments that are otherwise permissive for IAV fusion . To visualize single LASV entry and fusion , which is not restricted by IFITM3 in A549 cells [15] , we pseudotyped the HIV-1 core containing the bi-functional mCherry-2xCL-YFP-Vpr marker with the LASV GPc envelope glycoprotein complex to generate LASV pseudoparticles ( LASVpp ) . LASVpp imaging in A549 cells confirmed the ability to track single particles and detect their fusion ( release of mCherry ) in late endosomal compartments ( Fig 5B , S10 Movie ) . Interestingly , LASVpp fusion exhibited a unique feature rarely seen for other viruses , including IAV . The YFP-Vpr fluorescence , which is markedly decreased at mildly acidic pH [51 , 52] , was consistently quenched at some point prior to viral fusion , demonstrating acidification of intraviral pH [53 , 54] ( schematized in Fig 5A ) . Single frame images show that YFP-Vpr signal quenched for ~10 min before viral fusion , which is observed as the loss of mCherry signal ( red ) and concomitant reappearance of the YFP-Vpr signal ( Fig 5B and 5C , arrow ) . The dequenching of YFP fluorescence can be attributed to the re-neutralization of the virus’ interior through a fusion pore connecting it to the cytoplasm [52 , 54 , 55] . Based on the differences in IAVpp and LASVpp fusion , we classified single virus fusion events into “Type I” , in which mCherry signal is lost without acidification of the virus interior ( YFP quenching ) , as observed in IAV fusion , and “Type II” events , in which acidification of the virus interior occurs prior to fusion ( mCherry release ) , as observed in LASV fusion . In A549 cells , only 9% of LASVpp fusion events are Type I , while most particles—91%—undergo Type II fusion ( Fig 5D ) . Of note , YFP-Vpr quenching occurs for most particles not undergoing fusion at later times after infection due to a non-specific acidification of the viral interior in late acidic compartments . These events representing a non-productive entry of LASVpp were excluded from analysis . Additional experiments to confirm single LASVpp fusion in A549 cells were also performed . Control cells were treated with a broad-spectrum arenavirus entry inhibitor , ST-193 [56] , which abrogated single LASVpp fusion events ( Fig 5D ) . A total of 90 Type II events were observed in at least 24 independent experiments , with 6430 viral particles in Vector control cells and 5264 viral particles observed in cells treated with ST-193 . LASVpp fusion with A549 cells measured by a bulk BlaM assay also demonstrated potent inhibition of viral fusion in the presence of 10 μM ST-193 or upon raising the endosomal pH by 40 mM NH4Cl ( Fig 5E ) . Thus , the observed changes in fluorescent signals faithfully represent single LASVpp fusion . Consistent with the lag between YFP quenching and fusion ( Fig 5B and 5C ) , the kinetics of LASVpp fusion lagged behind the YFP quenching events ( Fig 5F ) . The average lag time between YFP-Vpr quenching and fusion for LASVpp in A549 cells is 14 . 1 minutes ( Fig 5F , inset ) . To test if the observed lag was due to the requirement for further virus trafficking to fusion-permissive compartments , we asked if it depended on how long a virus trafficked prior to YFP-quenching ( Fig 5G ) . The lag between quenching and fusion does not appear to be correlated with the waiting time for quenching ( R2 = 0 . 0263 ) , suggesting that LASVpp fusion following the YFP quenching is a stochastic event that does not depend on the virus trafficking history . Although acidification of the virus interior does not directly report the time of acidification of endosomal lumen , the above results demonstrate that LASV GPc retains fusion-competence under acidic conditions for a considerable time before it fuses with permissive late endosomes , perhaps after binding to LAMP1 [57 , 58] . We next assessed the basis for LASV resistance to IFITM3 restriction . A549-IFITM3-imTFP1 cells were infected with LASVpp labeled with mCherry-2xCL-YFP-Vpr , as above . Single particle imaging revealed that LASVpp did not co-traffic with IFITM3-imTFP1-positive endosomes and that subsequent viral fusion occurred at sites devoid of this fluorescent restriction factor ( Fig 6A , S11 Movie ) . Analysis of single LASVpp fluorescence intensities in A549-IFITM3-imTFP1 cells shows a typical Type II fusion event which occurs within an endosome lacking above-background amounts of IFITM3-imTFP1 ( Fig 6B ) . Live cell imaging experiments were performed at least 6 times independently , and on average , 1 . 64% and 1 . 87% of double-labeled LASVpp particles bound to cells fused in A549 Vector and A549-IFITM3-imTFP1 cells , respectively ( p = 0 . 551 ) ( Fig 6C ) . These data confirm previous reports that the expression of IFITM3 does not affect LASV fusion [15 , 30] . In addition , the kinetics of LASVpp fusion in control and A549-IFITM3-imTFP1 cells were not significantly different ( S6A Fig ) . LASVpp fusion kinetics were the same regardless of IFITM3-imTFP1 expression , as observed using the BlaM assay and stopping fusion at varied times by NH4Cl addition ( S6B Fig ) . We note that in one or two rare examples , LASVpp fusion appears to occur in an endosome containing detectable IFITM3-imTFP1 signal . Representative images and fluorescence traces show co-trafficking of a LASV particle within an IFITM3+ endosome until fusion occurs around 14 min post-infection ( Fig 6E and 6F and Inset , S12 Movie ) . This unique event is atypical of the majority of tracked particles due to several reasons: ( 1 ) there is an apparent colocalization with IFITM3+ beginning at time 0; and ( 2 ) LASVpp rarely fuse as early as 14 min post-infection . Most importantly , the fusion event in Fig 6E and 6F appears to represent transient fusion pore opening , as indicated by the sudden re-quenching of YFP , or re-acidification of the virus interior following pore closure ( Fig 6E–6G ) . We report this instance to illustrate that , while LASVpp typically avoids IFITM3+ endosomes , miniscule levels of fusion may occur within IFITM3+ compartments . Overall , LASVpp exhibit a strong tendency to bypass IFITM3+ endosomes and this important feature likely represents the mechanism by which this virus escapes restriction . Analysis of single IAVpp and LASVpp co-trafficking with IFITM3-imTFP1 endosomes in the context of fusion ( Figs 3–6 ) suggests that IAVpp would be trapped in IFITM3-positive endosomes/multivesicular bodies , while LASVpp would not . To test this notion , we followed the bulk virus uptake and transport in live A549 cells expressing IFITM3-imNG . Cells were incubated in the cold for 1 . 5 hr with IAVpp or LASVpp labeled with an internal fluorescent marker , Gag-mCherry , to allow virus binding . Cells were then incubated for indicated times ( 0 , 15 , 30 , and 60 min ) at 37°C , fixed and imaged at high spatial resolution . Representative images of IAVpp and LASVpp co-localization with IFITM3 containing vesicles at different time points are shown in Fig 7A . Quantification of virus colocalization with IFITM3-imNG over time shows that IAVpp increasingly co-localizes with IFITM3-imNG compartments , while LASVpp does not show a significant increase in colocalization up to 1 hr post-infection ( Fig 7B ) . These data support the hypothesis that restriction-sensitive viruses ( as is the case for IAV ) co-traffic with IFITM3 , while resistant viruses ( like LASV ) are transported through distinct endosomal compartments devoid of this restriction factor . To further test whether the presence of IFITM3 is necessary and sufficient to restrict IAV fusion , we generated control IAV particles that incorporated IFITM3 through co-expression in virus-producing cells . These pseudoviruses contained BlaM-Vpr to assess their fusion-competence . IFITM3 incorporation into virions and its possible effects on HIV-1 maturation or the influenza HA incorporation into pseudoviruses were verified by Western blotting ( Fig 8A ) . IFITM3 was present in pseudoviruses prepared in producer cells expressing IFITM3 but not in the Vector control . Furthermore , the amount of p24 protein and influenza HA were the same in the two preparations ( Fig 8A ) , indicating that IFITM3 incorporation does not perturb the expression or proteolytic processing of HA . To probe the fusion activity of IFITM3-containing pseudoviruses , A549 cells were incubated with IAVpp/IFITM3 ( or control viruses lacking IFITM3 ) at 4°C for 30 min , followed by incubation for 2 hours at 37°C in either drug-free medium or medium supplemented with AmphoB ( which rescues IAV fusion in A549-IFITM3-imTFP1 cells , Fig 4A and [35] ) . Compared to the control IAVpp , fusion of IAVpp containing IFITM3 was potently inhibited ( p<0 . 001 , Fig 8B ) . Interestingly , AmphoB rescued IAVpp/IFITM3 fusion ( Fig 8B ) , suggesting a direct effect of this antibiotic on IFITM3 or the viral membrane that is independent of cellular processes , including endocytic transport . The diminished ability of IAVpp produced in the presence of IFITM3 to fuse with target cells was not caused by IFITM3-containing extracellular vesicles present in the viral preparations , as have been suggested in [59] . Viral fusion was not significantly diminished when A549 cells were infected with a mixture of control IAVpp and extracellular medium from cells transfected only with an IFITM3-expressing plasmid ( S7A Fig ) . Although IFITM3-containig extracellular vesicles were effectively concentrated by our virus concentration protocol using LentiX ( S7B Fig ) , these vesicles did not modulate the fusion activity of IAVpp under our experimental conditions involving a brief exposure of target cells to the virus and vesicles . Taken together , these results suggest that the presence of IFITM3 is necessary and sufficient to restrict IAV fusion , irrespective of whether or not IFITM3 is expressed in the target or viral membrane . Further evidence supporting the proximity-based antiviral activity of IFITM3 was obtained by measuring LASV GPc-mediated cell-cell fusion . In this model , cell fusion is triggered by exposure to low pH , which bypasses the need for endocytic trafficking that may sort LASVpp away from IFITM3+ compartments . Cos7 cells transiently expressing LASV GPc were brought in contact with 293T cells stably expressing IFITM1 , IFITM2 or IFITM3 [22] or , in control experiments , with parental 293T cells . Cos7 and 293T cells were pre-loaded with different cytosolic fluorescent dyes to monitor fusion initiated by an acidic buffer , as previously described [22] . In stark contrast to LASVpp fusion with IFITM3-expressing cells ( Fig 6 ) , GPc-mediated cell-cell fusion was markedly inhibited by all three IFTIMs expressed in target cells ( Fig 8C ) . This result supports the notion that LASV GPc is not inherently resistant to IFITM restriction and that the reason LASV is insensitive to IFITM3 expression is through its usage of trafficking pathways that are distinct from those used by IFITM3 .
A remarkable breadth of enveloped viruses that are restricted by IFITM proteins suggests a universal mechanism for antiviral activity that likely involves altering the properties of the host cell membranes in a way that precludes viral fusion . It remains unknown how IFITMs exert their antiviral effects and , equally importantly , how arenaviruses and MLV escape restriction . In this study , we addressed a critical question of whether IFITMs work by a proximity mechanism , which requires their presence at the sites of virus entry and whether the lack of local IFITMs is a major determinant of virus resistance . Through constructing a functional fluorescently tagged IFITM3 protein , we were able to visualize its dynamic distribution in living cells , in the context of single virus entry and fusion . Imaging experiments demonstrate that IAV restriction involves virus co-trafficking with IFITM3-containing endosomes that can culminate in lipid mixing ( hemifusion ) but does not progress to complete fusion ( viral content release ) . This important finding , along with our previous work [17 , 30 , 35] , strongly supports a proximity model for virus restriction , as opposed to alternative models that involve , for example , dysregulation of cholesterol transport from late endosomes [18 , 39] . Also importantly , we documented the “avoidance” mechanism of LASVpp escape from IFITM3 restriction through virus trafficking and fusion with endosomes lacking this restriction factor . Consistently , LASV GPc-mediated cell-cell fusion is sensitive to IFITM proteins expressed on the surface of target cells . These results highlight the importance of regulation of IFITM trafficking for antiviral activity and offer important clues regarding the determinants of virus resistance to restriction . Of note , the presence of IFITM3 at the sites of IAV fusion does not rule out the possibility that the antiviral effect is due to recruitment of downstream effector proteins , such as ZMPSTE24 [60] . IFITMs have the propensity to hetero-oligomerize [21] and interact with a number of other proteins [61] , so it is possible that IFITM-driven protein complexes alter the membrane properties and disfavor viral fusion ( see below ) . Single particle tracking revealed that IAV fusion was inhibited in compartments that accumulated substantial amounts of fluorescent IFITM3 . Due to the relatively high and variable background fluorescence in cells expressing fluorescent IFITM3 , it is difficult to quantitatively assess whether there is a threshold density of this protein below which viruses are not restricted . In other words , it is unclear whether inhibition of IAV fusion by IFITM3 occurs through an all-or-none mechanism or there is an inhibition “gradient” whereby the probability of fusion is inversely proportional to the IFITM3 signal . Future studies using improved fluorescence labeling techniques and controlled IFITM3 expression levels will help distinguish between these modes of action . The existence of distinct domains within the highly dynamic endosomal membranes ( e . g . , [62–66] ) adds an additional layer of complexity when interpreting the IFITM3 restriction results . It is possible that the extremely rare single LASVpp fusion events that appear to colocalize with IFITM3 occur with IFITM3-free domains within the limiting membrane of an endosome . We have previously documented unimpeded IAV lipid mixing activity in IFITM3-expressing cells [30] . Analyses of lipid dye dequenching , irrespective of colocalization with IFITM3 , did not reveal significant differences in the rate or extent of lipid mixing . This finding is in disagreement with the reduced IAV lipid dequenching in IFITM3-expressing cells reported in [39] . The reason for discrepant results is likely related to the use of a bulk lipid dequenching assay in [39] , as compared to the real-time single IAV tracking in our experiments . Importantly , in the present study , we were able to show the markedly slower rate of lipid redistribution to IFITM3-containing endosomes by focusing on events occurring in these compartments compared to lipid mixing in control cells . It should be noted that , in spite of exogenous incorporation of DiI into the viral membrane in the commonly used labeling protocol ( e . g . , [30 , 67] ) , the dye readily redistributes to both membrane leaflets , as we have demonstrated previously [68] . Thus , a slower lipid mixing between IAV and IFITM3-positive endosomes is consistent with a more restricted dye diffusion through the merged contacting leaflets of hemifused membranes , as compared to diffusion through both leaflets of a fusion pore . The exact mechanism by which IFITM3 inhibits the transition from hemifusion to fusion is not clear . A large body of work demonstrates a critical role of lipid composition , and specifically of mechanical properties of lipid membranes , in protein-mediated membrane fusion ( reviewed in [69] ) . Bending energies of highly curved lipid intermediates that form and resolve during merger of lipid bilayers are key determinants of the fusion process ( reviewed in [69–71] ) . In addition , hemifusion and the formation of a fusion pore within a hemifusion diaphragm are associated with changes in areas of contacting and distal monolayers , respectively . Thus , viral fusion pore opening could be blocked by: ( 1 ) increased membrane bending modulus; ( 2 ) increased negative curvature of the cytoplasmic leaflet that disfavors the formation of a net positive curvature fusion pore [69 , 72]; ( 3 ) expansion of the hemifusion diaphragm to a size beyond that permissible for fusion pore formation [73]; or ( 4 ) reduced “fluidity” ( lateral diffusion ) of the cytoplasmic leaflet , which can be caused by IFITM homo/hetero-oligomerization [21] . The latter effect is expected to disfavor the fusion pore opening due to inability to quickly remove excess lipid from the hemifusion site . IFITMs have been reported to alter membrane fluidity [21 , 35] , and to increase the lipid order and confer positive spontaneous curvature [22 , 33] . It is thus possible that individual effects of IFITMs on lipid membranes or their combination are responsible for the fusion block . Importantly , a recent study demonstrated that mutations in distinct regions of IFITM3 regulate its inhibitory vs enhancing activity against infection by different coronaviruses [74] . The ability to switch between inhibition and promotion of coronavirus fusion by introducing point mutations in IFITM3 further supports the proximity-based mechanism of virus restriction . We have previously proposed an alternative mechanism of IFITM3-mediated virus restriction referred to as a “fusion decoy” model [30] . According to this model , viruses are redirected into multivesicular endosomes where unrestricted fusion with intraluminal vesicles , as opposed to fusion with the limiting membrane of an endosome , does not allow viral capsid release into the cytoplasm . The single virus content ( mCherry ) release assay would not detect IAVpp fusion with intraluminal vesicles , as the content marker will remain contained within the same endosome . The similar extent of lipid dye dequenching upon single IAV fusion with control and IFITM3-positive endosomes ( Fig 2 ) appears compatible with virus hemifusion to the limiting membrane , but the slower dequenching rate in IFITM3 compartments could be due to multiple rounds of hemifusion with intraluminal vesicles . Therefore , additional experiments are needed to test the validity of a “fusion decoy” model . Recent studies have documented the ability of IFITMs to interfere with viral fusion when incorporated into the viral membrane [34 , 36–38] . In fact , IFITMs appear to more potently inhibit HIV-1 infection when incorporated into virions , as compared to their expression in target cells [38] . It is tempting to assume that the same mechanism of the IFITMs’ antiviral activity is functional in both cellular and viral membranes , but this notion has not been explicitly tested . The ability of IFITM3 to inhibit IAV fusion irrespective of whether it is expressed in the target or viral membrane ( Fig 8B ) supports the universal mechanism of IFITM3-mediated restriction that involves altering the properties of lipid membranes , as opposed to interacting with viral or cellular proteins . Our finding that AmphoB rescues the fusion-competence of IAVpp containing IFITM3 , similar to its antagonistic effect on the cell-expressed IFITM3 [35] , is also consistent with the common mechanism of virus restriction . Moreover , the static nature of the viral membrane , which is in stark contrast to the highly dynamic cell membranes , supports a direct effect of AmphoB on the viral membrane , perhaps through alterations of membrane fluidity [22 , 35] . Thus , virions containing IFITMs in their membranes could provide a tractable model for mechanistic studies of these proteins . Our study focused on IFITM3 protein , which shares a relatively high sequence homology and subcellular distribution with IFITM2 . Although we have not addressed the mechanism of action of the plasma membrane-resident IFITM1 , the published literature and our findings support the notion that this protein also acts by a proximity-based mechanism . We thus speculate that all members of the IFITM family accumulate at the sites of fusion of sensitive viruses and block the formation of a fusion pore . The current study provided strong evidence that LASV escapes IFITM3 restriction by entering through alternative endocytic pathways , but has not addressed whether other IFITM-resistant viruses , such as Junin virus or MLV , employ the same strategy to infect IFITM-expressing cells . Future studies addressing this question will help generalize the escape mechanism discovered in this work and may suggest strategies to increase the potency of IFITMs by modulating their intracellular trafficking .
We obtained HEK 293T/17 , Cos7 and human lung epithelial A549 cells from ATCC ( Manassas , VA ) . TZM-bl cells were obtained from NIH AIDS Research and Reference Reagent Program . 293T cells stably expressing IFITM1 , IFITM2 or IFITM3 were a gift from Dr . Shan-Lu Liu , Ohio State University [22] ) . Cells were maintained in Dulbecco’s Modified Eagle Medium ( DMEM , Cellgro , Mediatech , Masassas , VA ) containing 10% heat-inactivated Fetal Bovine Serum ( Hyclone Laboratories , Logan , UT or Atlanta Biologicals , Flowery Branch , GA ) and 1% penicillin/streptomycin from Gemini Bio-products ( West Sacramento , CA ) . For HEK 293T/17 cells the growth medium was supplemented with 0 . 5 mg/ml G418 ( Genesee Scientific , San Diego , CA ) . DMEM without phenol red was purchased from Life Technologies ( Grand Island , NY ) . Live Cell Imaging Buffer ( LCIB ) and FluorobriteTM DMEM were purchased from Life Technologies ( Grand Island , NY ) . Stable cell lines expressing fluorescently-tagged IFITM3 used for imaging analysis were obtained by transducing with VSV-G pseudotyped viruses encoding wild-type or the 2M mutant ( F75/78A ) IFITM3 [21] or with the Vector pQCXIN ( Clontech ) and selecting with 1 mg/ml G418 . Following selection , cells were maintained in 0 . 5 mg/ml G418 . Stable cell lines expressing unlabeled and fluorescently-tagged IFITM3 used for bulk fusion assays were obtained by transducing with VSV-G pseudotyped viruses encoding wild-type IFITM3 with the Vector pQXCIP ( Clontech ) and selecting with 1 . 5 μg/ml puromycin . The pR8ΔEnv , pR9ΔEnv , BlaM-Vpr , pcRev , pMDG-VSV-G , MLV-Gag-Pol , HIV-1 Gag-mCherry ( encoding an uncleavable mCherry fluorescent tag used in fixed cell co-localization experiments ) , IFITM3/pQCXIP , F75/78A IFITM3/pQCXIN expression vectors were described previously [15 , 21 , 43 , 75] . The mCherry-2xCL-YFP-Vpr ( mCherry fused to YFP-Vpr through a cleavable linker containing two HIV protease cleavage sites– 2xCL ) , as described previously [50] , was used for single particle tracking of fusion events in live cells . The pCAGGS vectors encoding influenza H1N1 WSN HA and NA were provided by Drs . Donna Tscherne and Peter Palese ( Icahn School of Medicine , Mount Sinai ) [76] . The LASV GPc plasmid was a gift from Dr . F . -L . Cossett ( Université de Lyon , France ) [77] . To internally label human IFITM3 ( accession NM_021034 ) with EGFP , sites likely permissive to insertions were identified based on sequence alignments of human and mouse IFITM family proteins . An EGFP cassette flanked by two linkers was created by PCR ( forward primer TCAAGGAGGAGCACGAGGTGGCTGTGCTGGGGGCGCCCCACAACCCTGCTCCCGGCGGAGGAAGCGGCGGAGTGAGCAAGGGCGAGGAGC; reverse primer GACGACATGGTCGGGCACGGAGGTCTCGCTGCGGATGTGGATCACGGTGGATCCGCCTCCGCTTCCGCCCTTGTACAGCTCGTCCATGCC ) and inserted using Gibson assembly into KasI/BsaBI-cut IFITM3 cDNA . The resulting protein has amino acids 41 ( P ) and 42 ( T ) of wild-type IFITM3 removed . The cloning of IFITM3-imNeonGreen ( IFITM3-imNG ) , and IFITM3-imTFP1 ( IFITM3-imTFP1 ) into pQCXIN ( Clontech ) and pQXCIP ( Clontech ) retroviral expression vectors were done in two steps . In the first step , the EGFP in IFITM3-iEGFP/pLVX Tet on construct was replaced either with mNeonGreen or mTFP1 by overlapping PCR . The IFITM3-iEGFP/pLVX Tet on construct contain an EcoRI restriction site in the 5’ of the IFITM3-iEGFP cDNA and a BsrgGI in the 3’ of GFP cDNA that facilitated the replacement of GFP . The 5’ of IFITM3 cDNA was amplified by PCR using forward primer P1 ( containing EcoRI restriction site ) TACCACTTCCTACCCTCGTAAAGAATTCGCCACCATGAATCACACTGTCCAAACCTTC , and reverse primer P2: TGTGGTCTCCTCGCCCTTGCTCACTCCGCCGCTTCCTCCGCCGGGAGC . The mTFP1 fragment was amplified using forward primer P3: GCTCCCGGCGGAGGAAGCGGCGGAGTGAGCAAGGGCGAGGAGACCACA ( complementary to P2 ) , and the reverse primer P3 ( containing BsrgGI restriction site ) GATCCGCCTCCGCTTCCGCCCTTGTACAGCTCGTCCATGCCGTCGGTGGAATT . The fragments were purified , mixed and the overlapping PCR was perfomed using the forward primer P1 and the reverse primer P3 . The final PCR fragment and IFITM3-iEGFP/pLVX Tet on were digested with EcoRI and BsrgGI restriction enzymes , purified and ligated . In the second step , the IFITM3-imTFP was amplified by PCR with forward primer P4 ( containing AgeI restriction site ) GCAGGAATTGATCCGCGGCCGCACCGGTAGGCCACCATGAATCACACTGTCCAAACCTTC , and reverse primer P5 ( containing EcoRI restriction site ) AGGGGTGGGGCGGGGGGGGGCGGAATTCTTAGTGATGGTGATGGTGATGGCCTTG , digested with AgeI and EcoRI restriction enzymes , purified and ligated into pQCXIN or pQXCIP vectors . For IFITM3-imNeonGreen construct the overlapping PCR was done using IFITM3-imTFP/pQCXIN construct as template and the AgeI and BamHI restriction sites . The 5’ of IFITM3 cDNA was amplified by PCR using forward primer P5 ( containing AgeI restriction site ) GCAGGAATTGATCCGCGGCCGCACCGGTAGGCCACCATGAATCACACTGTCCAAACCTT , and reverse primer P6: ATCCTCCTCGCCCTTGCTCACCATTCCGCCGCTTCCTCCGCCGGGAGC . The mNeonGreen cDNA was amplified with forward primer P7: GCTCCCGGCGGAGGAAGCGGCGGAATGGTGAGCAAGGGCGAGGAGGAT ( complementary to P6 ) , and reverse primer ( containing BamHI restriction site ) CGCTGCGGATGTGGATCACGGTGGATCCGCCTCCGCTTCCGCCCTTGTACAGCTCGTCCATGCCCA . Purified PCR fragments were mixed and the overlapping PCR was done using the forward primer containing AgeI restriction site and the reverse primer containing BamHI restriction site . The fragment and IFITM3-imTFP1 plasmid were digested with AgeI and BamHI , purified and ligated . The F75/F78A IFITM3-imTFP1 ( 2M-IFITM3-imTFP1 ) mutants was obtained by Quick-change site-directed mutagenesis ( Stratagen , La Jolla , CA ) using IFITM3-imTFP1/pQCXIN as template . Alexa Fluor 647-NHS ester ( AF647 ) and the lipophilic dye SP-DiIC18 ( 1 , 1'-Dioctadecyl-6 , 6'-Di ( 4-Sulfophenyl ) -3 , 3 , 3' , 3'-Tetramethylindocarbocyanine ) were purchased from Invitrogen/Life Technologies ( Grand Island , NY ) . The LASV fusion inhibitor ST-193 was purchased from Aurum Pharmatech ( Franklin Park , NJ ) . Amphotericin B ( AmphoB ) was obtained from Quality Biological ( Gaithersburg , MD ) . Primary antibodies used were rabbit directed at the N-terminus of IFITM3 ( Abgent , San Diego , CA ) , mouse anti-tubulin from Sigma ( St . Louis , MO ) , HIV-1 IG serum ( NIH AIDS Research and Reference Reagent Program ) , and rabbit anti-WSN Influenza R2376 ( a generous gift from Dr . David Steinhauer , Emory University ) . Secondary antibodies used were rabbit anti-mouse IgG ( H+L ) -HRP ( EMD Millipore ) , mouse anti-rabbit IgG ( H+L ) -HRP ( EMD Millipore ) , and goat anti-human IgG-HRP ( H+4 ) ( Thermo Scientific ) . Pseudoviruses were produced by transfecting HEK 293T/17 cells with JetPRIME transfection reagent ( Polyplus-transfection , Illkirch-Graffenstaden , France ) . For all pseudovirus productions the transfection reagent/DNA containing medium was replaced with fresh phenol red-free medium after ~14 hrs . Viruses were harvested ~48 hrs post-transfection , and cellular debris were removed by centrifuging at 230xg for 10 min . The collected viruses were passed through a 0 . 45 μm polyethersulfone filter ( PES , VWR ) to further clear cellular debris and virus aggregates , aliquoted and stored at -80°C . The infectious titers ( ~106 IU/ml ) were determined using serial dilutions of the inoculum in TZM-bl cells using a β-galactosidase assay . To produce the pseudoviruses for co-localization analysis , HEK293T/17 cells were grown to ~60–70% confluency in a 6-well culture dish and transfected with 0 . 8 μg pR8ΔEnv , 0 . 4 μg pcRev , 0 . 5 μg HIV-1 Gag-mCherry ( Not cleaved by protease ) and 0 . 8 μg GPc-Lassa or 0 . 4 μg each of WSN HA- and NA-expressing plasmids , respectively . For single viral fusion experiments in live cells , dual-labeled LASVpp was made by transfecting the HEK293T/17 cells grown to ~60–70% confluency in a 6-well culture dish using 0 . 8 μg pR9ΔEnv , 0 . 2 μg pcRev , 0 . 2 μg mCherry-2pxCLYFP-Vpr and 1 μg GPc-Lassa plasmids . Similarly , dual-labeled IAVpp were produced using 4 μg pR9ΔEnv , 1 μg pcRev , 1 μg mCherry-2xCL-YFP-Vpr and 2 . 5 μg each of WSN HA- and NA-expressing plasmids for transfection of ~60% confluent cells in a 100 mm dish . The purified viruses were diluted 10-fold in PBS without calcium or magnesium ( PBS-/- , Cellgro , Mediatech ) , bound to poly-L-lysine coated 8-well chamber cover slips ( LabTek , MA ) , and imaged to estimate the co-labeling efficiency which was over 90% for all pseudoviruses used for this study . To generate VSV-G pseudotyped viruses encoding fluorescently-tagged IFITM3 , HEK293T/17 cells grown in 6-well plate were transfected with 0 . 3 μg of VSV-G plasmid , 0 . 6 μg MLV-Gag-Pol plasmid , and 1 . 1 μg of either an empty pQXCIN or pQXCIP vector , or containing IFITM3-imNG , IFITM3-imTFP1 , or 2M-IFITM3-imTFP1 . For intraviral IFITM3 pseudovirus production , HEK293T/17 cells grown in 100-mm dishes were transfected with 2 μg of pCAGGS H1N1 HA/NA , 3 μg pR9ΔEnv , 1 . 5 μg BlaM-Vpr , 0 . 5 μg pcRev , and 5 μg of either empty Vector pQCXIP , IFITM3 , or 2M-IFITM3 using JetPRIME reagent . The viral supernatants cleared of cellular debris as described above , were concentrated 10x , using Lenti-X Concentrator ( Clontech , Mountain View , CA ) . Following overnight concentration with Lenti-X , virus was precipitated by centrifuging at 1439xg for 45 min , 4°C , resuspended in DMEM without phenol red or FBS , and stored at -80°C . For lipid mixing ( hemifusion ) experiments , influenza virus surface proteins and membrane were co-labeled with AF647 and with the lipophilic dye SP-DiIC18 , respectively . Briefly , 100 μg of the purified IAV A/PR/8/34 virus ( 2 mg/ml , Charles River , CT ) was mixed with 50 μM AF647 in 150 mM freshly prepared sodium bicarbonate buffer , pH 9 . 0 . The labeling reaction was allowed to proceed at room temperature with tumbling in the dark for 30 min . Next , 5 . 8 μL of 1 . 75 mM SP-DiI18 was added to this reaction , while gently vortexing and viruses further incubated at room temperature in the dark for 1 hr with shaking . The AF647 was quenched by adding 2 μL of 1 M Tris-buffer , pH 7 . 0 . The labeled viruses were purified from excess dyes on a Nap-5 gel filtration column ( GE Healthcare ) that was equilibrated with 50 mM HEPES , pH 7 . 4 , 145 mM NaCl at room temperature . The fractions containing labeled viruses were passed through a 0 . 45 μm filter to remove any large lipid and/or virus aggregates . The purified viruses were bound to poly-L-lysine coverslips and imaged to quantify their co-labeling efficiency which was at least 55% , determined as the percentage of AF647 labeled viruses that showed detectable signal ( under the high self-quenching concentrations ) of SP-DiI18 . The viruses were aliquoted into tubes , flash-frozen , and stored at -80°C until use . The β-lactamase ( BlaM ) assay for virus-cell fusion were performed as described previously [30 , 43] . Briefly , pseudoviruses containing a β-lactamase-Vpr chimera ( BlaM-Vpr ) were bound to target cells by centrifugation at 4°C for 30 min at 1550xg . Unbound viruses were removed by washing with DMEM without phenol red supplemented with 20 mM HEPES ( GE Healthcare Life Sciences ) . Fusion was initiated by shifting to 37°oC for 2 hours , after which cells were placed on ice and loaded with the CCF4-AM substrate ( Life Technologies ) and incubated overnight at 11°C . The cytoplasmic BlaM activity ( ratio of blue to green fluorescence ) was measured using a SpectraMaxi3 fluorescence plate reader ( Molecular Devices , Sunnyvale , CA ) . The p24 content of viral stocks was determined by ELISA , as described previously [78] . Whole cell lysates were harvested in RIPA Buffer ( Sigma ) supplemented with protease inhibitors ( Complete Protease Inhibitor Cocktail , Roche ) , incubated on ice for 10 min , and cleared by centrifugation at 16 , 000xg for 5 min . Total protein was measured using a bicinchoninic acid assay ( BCA , Pierce ) and normalized protein was loaded onto 4–15% polyacrylamide gels ( Bio-Rad , Hercules , CA ) . Precision Plus Protein Standards ( Kaleidoscope Bio-Rad ) were used as molecular weight markers . Proteins were transferred onto a nitrocellulose membrane , blocked in 10% Blotting-grade Blocker ( Bio-Rad ) in PBS-T ( phosphate buffered saline with 0 . 1% Tween-20 ) for 30 min at room temperature . Membranes were incubated in primary antibodies overnight at 4°C in 5% Blotting-grade Blocker with gentle shaking: rabbit anti-IFITM3 ( 1:500 ) , mouse anti-tubulin ( 1:3000 ) , human HIV-IG ) ( 1:2000 ) , and rabbit anti-WSN Influenza R2376 ( 1:100 ) . After washing membranes with PBS-T at room temperature , Horseradish peroxidase-conjugated ( HRP ) goat anti-rabbit , rabbit anti-mouse , and goat anti-human secondary antibodies were added in 5% Blotting-grade Buffer for 1 h at room temperature with gentle shaking . Following PBS-T washing of membranes , ECL Prime chemiluminescence reagent ( GE Healthcare ) was used for protein detection . The effector Cos7 cells were transfected with the Lassa virus GPc expression vector . Briefly , cells were grown on 35 mm culture dishes to ~60% confluency and transfected with 4 μg GPc expression vector using a calcium-phosphate protocol [22] . After 48 hours following transfection , cells were loaded with 1 . 3 μM of the green cytoplasmic Calcein-AM dye ( Invitrogen ) . In parallel , 293T cells or their derivatives stably expressing human IFITM1 , IFITM2 or IFITM3 [22] were labeled with 30 μM of the blue cytoplasmic dye CMAC ( Invitrogen ) . Effector and target cells were washed , detached from the culture dishes using a non-enzymatic solution , resuspended in PBS++ , mixed at a 1:1 ratio and co-plated onto 8-well chamber slides . After incubating for 30 min at room temperature , cells were exposed to a pH 5 . 0 buffer at 37°C for 20 min , and the resulting cell-cell fusion was measured by visual inspection under a fluorescent microscope , as described in [22] . Ten fields of view each containing 10–12 heterologous cell pairs were examined in each well . A549 cells expressing IFITM3-imNG were cultured on collagen-coated 8-chamber coverslips ( Lab-Tek , NY #1 . 5 glass ) in Fluorobrite DMEM to ~60–80% confluency . The cells were chilled by placing on ice for 10 min , followed by aspirating the media and washing with cold phosphate buffered saline with calcium and magnesium ( PBS+/+ ) . Cells were inoculated with a 5-fold dilution of the mCherry-labeled IAV or LASV pseudoviruses in 100 μL of cold LCIB supplemented with 2% FBS , and viruses were allowed to bind to cells by incubation at 4°C for 90 min . Unbound viruses were removed by washing with cold PBS+/+ , and virus entry was initiated by adding 200 μL pre-warmed ( 37°C ) LCIB . The slides were incubated at 37°C for varied times followed by fixation with 4% paraformaldehyde ( PFA ) in PBS-/- for 10 min at 37°C . For the zero time-point , cells were fixed with PFA immediately following the initial virus binding step at 4°C . After fixation , PFA was washed away with PBS-/- a few times and cells were imaged . For co-localization analysis , the fixed cells were imaged on DeltaVision Elite ( GE Healthcare ) widefield microscope , using an Olympus PlanApoN 60x/1 . 42 NA oil immersion objective . Multiple Z-stacks with a spacing of 0 . 1 μm covering the entire thickness of the cells were acquired using a standard GFP/Cherry filter set . Deconvolution of the Z-stacks was done post acquisition to improve the signal-to-background ratio of the IFITM3-mNeonGreen vesicles and of mCherry labeled viruses using SoftWorX ( DeltaVision , GE Healthcare ) . The deconvolved Z-stacks were used for quantitative volume based ( voxel based ) co-localization analysis using a custom protocol in the image analysis program Volocity ( Perkin Elmer , Waltham , MA ) . Briefly , after background subtraction , the IFITM3 containing vesicles were identified as objects . mCherry virus particles with a size/volume threshold of 0 . 512 μm3 ( corresponding to a 2x2x2 voxel ) and that were associated with IFITM3-expressing cells were identified . A virus particle was considered co-localized with IFITM3 when at least 50% of its volume overlapped within an IFITM3 expressing object . For every time point , at least 4 different fields of view containing multiple cells were imaged and the mean values of the % co-localization calculated . A549 IFITM3-imNG , IFITM3-imTFP1 , or 2MIFITM3-imTFP1 cells were seeded onto 35 mm collagen-coated glass-bottom Petri dishes ( MatTek , MA ) one day prior to imaging and cultured in DMEM without phenol red supplemented with 10% FBS , penicillin , and streptomycin . Following a wash with room temperature PBS , cells were fixed in 3 . 5% paraformaldehyde in PBS for 10 min . For imaging with LysoTracker™ , cells were seeded as above and incubated with 30 nM LysoTracker™ Red DND-99 ( ThermoFisher ) diluted in pre-warmed LCIB supplemented with 2% FBS for 30 min prior to fixation . Images were acquired on a DeltaVision microscope using an Olympus UPlanFluo 40x/1 . 3 NA oil immersion objective ( Olympus , Japan ) . Multiple Z-stacks with a spacing of 0 . 1 μm covering the entire thickness of the cells were acquired and deconvolved . A549 cells were seeded onto 35 mm collagen-coated glass-bottom Petri dishes ( MatTek , MA ) one day before imaging and cultured in Fluorobrite DMEM supplemented with 10% FBS , penicillin , streptomycin and L-glutamine . Before imaging , the cells were pre-chilled on ice and washed with ice-cold PBS+/+ . A small amount of viral suspension ( ~1 μL ) diluted in 60 μL of cold LCIB was added to the cells and spinoculated at 1550xg at 4°C for 20 min . After spinoculation , the cells were washed twice with PBS +/+ to remove any unbound viruses and a small volume ( ~150 μL ) of ice-cold LCIB was added to the cells . Viral entry was initiated by adding 2 mL of pre-warmed ( 37°C ) LCIB containing 2% FBS , and cells were imaged immediately on DeltaVision microscope equipped with a temperature and humidity-controlled chamber . Every 6–8 sec , at least three Z-stacks spaced by 1 . 5–2 μm were acquired to cover the thickness of cells using Olympus 40x UPlanFluo 40x/1 . 3 NA oil objective ( Olympus , Japan ) . The three-color viral fusion experiments were done with A549 cells expressing IFITM3-imTFP1 or 2M-imTFP1 using a standard CFP/YFP/Cherry filter set ( Chroma , VT ) , while a TRITC/FITC/Cy-5 filter set was used for the three-color lipid mixing experiments with IFITM3-mNeonGreen expressing cells . LCIB in all the experiments was supplemented with 2% FBS . The time-lapse Z-stack movies are visually inspected as maximum intensity projections , using ImageJ , and the ROI manager tool was used to annotate the single fusion or hemifusion events ( observed as color change or DiI intensity increase ) . The sets of waiting times for hemi/fusion obtained from multiple movies are combined from experiments done on different days , sorted and plotted as cumulative probability curves that show the kinetics of the event . Acquired image series were converted to maximum intensity projections and annotated particles were tracked using either Volocity ( GE Healthcare ) or ICY image analysis software ( icy . bioimageanalysis . org ) . Fluorescently labeled viral particles were identified using the spot detection algorithm and tracked in 3D to determine fluorescence intensities at every time point . With three-color imaging to track double-labeled viral particles that co-traffic with fluorescently-labeled IFITM3 compartments , single Z-planes in which the viral particle trafficked were used so that background subtraction could be performed using the ICY spot tracking algorithm . The local background was determined by dilating the identified objects corresponding to viral particles by two pixels . The difference between the integrated intensities of the particle and the dilated surrounding gave the intensity surrounding the particle , from which an average per pixel local background was calculated . This was used to obtain the background-corrected intensity of the particle at every time point , which is plotted in the time traces as shown in the figures . For the lipid mixing experiments , SP-DiI18 dequenching curves for the individual viruses were obtained by tracking particles , either using the AF647 channel or the DiI channel . The dequenching ratios and times were manually obtained from time traces that show at least 4-fold increase in SP-DiI18 intensity . The dequenching ratio was calculated as ratio of the mean SP-DiI18 intensity before the rise of the signal and the mean intensity after completion of dequenching . The dequenching time was measured as time taken to reach an intensity plateau from the time of initial rise in intensity . The traces showing multi-phase increase in SP-DiI18 signal in IFITM3 expressing cells were excluded from the calculation of dequenching ratios and times .
|
Expression of interferon-induced transmembrane proteins ( IFITMs ) in target cells potently inhibits fusion of many unrelated enveloped viruses , including the Influenza A virus , whereas arenaviruses , such as the Lassa fever virus , are resistant to these factors . The mechanism by which IFITMs interfere with the viral fusion step and the mechanism of virus escape from these restriction factors are poorly understood . Here , we tagged the late endosome-resident IFITM3 with fluorescent proteins and visualized single virus entry and fusion with endosomes in living cells expressing these constructs . Single virus and endosome tracking experiments demonstrate that the sensitive Influenza A virus is trapped within acidic IFITM3-positive endosomes that are not permissive for viral fusion . In contrast , the resistant Lassa virus consistently enters and fuses with endosomes lacking IFITM3 . Our results imply that accumulation of IFITM3 in virus-carrying endosomes is a prerequisite for blocking fusion of diverse enveloped viruses and that viruses insensitive to this protein escape restriction by entering through distinct endosomal trafficking pathways that do not converge with IFITM3-positive compartments .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"physiology",
"fluorescence",
"imaging",
"medicine",
"and",
"health",
"sciences",
"vesicles",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"pathogens",
"microbiology",
"orthomyxoviruses",
"viruses",
"membrane",
"fusion",
"rna",
"viruses",
"cellular",
"structures",
"and",
"organelles",
"endosomes",
"research",
"and",
"analysis",
"methods",
"influenza",
"a",
"virus",
"lipids",
"viral",
"entry",
"imaging",
"techniques",
"medical",
"microbiology",
"microbial",
"pathogens",
"cell",
"membranes",
"biochemistry",
"cell",
"biology",
"influenza",
"viruses",
"virology",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"organisms",
"cell",
"fusion"
] |
2019
|
Interferon-induced transmembrane protein 3 blocks fusion of sensitive but not resistant viruses by partitioning into virus-carrying endosomes
|
Many scarab beetles have sexually dimorphic exaggerated horns that are an evolutionary novelty . Since the shape , number , size , and location of horns are highly diverged within Scarabaeidae , beetle horns are an attractive model for studying the evolution of sexually dimorphic and novel traits . In beetles including the Japanese rhinoceros beetle Trypoxylus dichotomus , the sex differentiation gene doublesex ( dsx ) plays a crucial role in sexually dimorphic horn formation during larval-pupal development . However , knowledge of when and how dsx drives the gene regulatory network ( GRN ) for horn formation to form sexually dimorphic horns during development remains elusive . To address this issue , we identified a Trypoxylus-ortholog of the sex determination gene , transformer ( tra ) , that regulates sex-specific splicing of the dsx pre-mRNA , and whose loss of function results in sex transformation . By knocking down tra function at multiple developmental timepoints during larval-pupal development , we estimated the onset when the sex-specific GRN for horn formation is driven . In addition , we also revealed that dsx regulates different aspects of morphogenetic activities during the prepupal and pupal developmental stages to form appropriate morphologies of pupal head and thoracic horn primordia as well as those of adult horns . Based on these findings , we discuss the evolutionary developmental background of sexually dimorphic trait growth in horned beetles .
Beetle horns are used as weapons for intraspecific combats between males . Beetle horns display sexual dimorphism in many Scarab species , and their shapes , numbers , sizes and forming regions are highly diverged even among closely related species [1–3] . Interestingly , the diversified horn forms are associated with the fighting styles employed by the beetles , such as scooping up , piercing and throwing [4] . Furthermore , beetle horns are thought to be an evolutionary novelty . Horns are an outgrowth structure derived not from an appendage but from a dorsal epidermal sheet . Elucidating how these novel traits were acquired in Scarab species will lead to better understanding the mechanisms of morphological diversification during evolution . Therefore , beetle horns are an attractive model for studying not only the association of trait novelty with sexually dimorphic development but also the evolution of novel traits . Beetle horn development has been investigated in several horned beetles including the Japanese rhinoceros beetle , Trypoxylus dichotomus ( Coleoptera , Scarabaeidae , Dynastinae ) . In T . dichotomus , male adults have sexually dimorphic exaggerated horns on the head and prothorax , which are used in combat among conspecific males as weapons [5 , 6] . The head horn is shaped like a plow with a long stalk , and bifurcated twice at the distal tip , while the prothoracic horn is shorter than the head horn , and bifurcated once at the distal tip . During development , the horns are first formed as thickened epidermal primordia at the prepupal stage . Male adults have exaggerated horns at the head and prothoracic regions , whereas females do not have these structures . However , females have a small head horn at the pupal stage ( Fig 1B , S1 Fig ) . Sexual dimorphism of horns can be first observed in horn primordia formed during the prepupal stage , and the horn primordia grow larger through cell growth before pupation [7 , 8] . In males , the length of the pupal horn primordia after pupation is almost the same as that of an adult horn [9] . However , the shape of a pupal horn is rounded , and slightly larger than an adult horn ( Fig 1B , S1 Fig ) . During the pupal period , the horn primodium is transformed to become an sophisticated adult horn shape through a process known as “horn remodeling” during which programmed cell death sculpts specific regions of the horn primordium into the adult morphology [7 , 10 , 11] . In insects , orthologs of a transcription factor gene doublesex ( dsx ) play a pivotal role in sexual differentiation . These orthologs are involved in the formation of sexually dimorphic traits through the expression of sex-specific isoforms ( dsxM and dsxF ) [8 , 12–17] . In addition , tissue-specific expression of dsx is elevated in regions where sexually dimorphic structures are formed to induce their development [17–21] . In horned beetles including T . dichotomus , the dsx orthologs also regulate sexually dimorphic horn formation during larval-pupal development . RNA interference ( RNAi ) targeting dsx results in intersexual phenotypes both in males and females [8 , 14] . The molecular mechanisms of sex-specific splicing of dsx has been intensely studied in Drosophila melanogaster [12 , 22] . In D . melanogaster , Sex-lethal ( Sxl ) , the master sex determination gene , initiates the sex determination cascade . Sxl encodes an RNA-binding protein that directly binds to target RNAs . Functional Sxl protein is translated only in females [23–25] , and controls sex-specific splicing of transformer ( tra ) to produce functional Tra protein in females [26] . This TRA molecule then forms a heterodimer with a ubiquitously expressed RNA-binding protein , Transformer2 ( Tra2 ) , to regulate the sex-specific alternative splicing of dsx by directly binding to dsx transcripts in females [27 , 28] . The resultant dsxF isoform in females and dsxM isoform in males regulate a battery of downstream genes to form sexually dimorphic traits by binding to the target DNA sequences with the Dsx binding motif [29 , 30] , and by activating or repressing their transcription [31] . intersex ( ix ) functions as a female-specific co-activator by directly binding to female-specific isoforms of the DsxF protein , and regulates development of female-specific traits [32] . The Sxl-[tra/tra2]-[dsxF/ix] pathway described above is only activated in females . In males , default mRNA splicing results in expression of the male-specific splicing variant of dsx , dsxM [12 , 22] . Loss of function of the sex determination genes in females , and gain of function in males result in sex transformation [12 , 22] . Orthologous genes corresponding to the D . melanogaster sex determination genes described above are conserved among many holometabolous insects . In addition , the function of splicing regulatory factors ( Tra/Tra2 ) and a transcription co-factor ( Ix ) that are supposed to directly interact with dsx transcripts or Dsx protein are conserved in many holometabolous insects ( e . g . Diptera: Drosophila melanogaster [33] , the housefly Musca domestica [34] , the Mediterranean fruit fly Ceratitis capitata [35] , the Australian sheep blowfly Lucilia cuprina [36] , the oriental fruit fly Bactrocera dorsalis [37] , Hymenoptera: the honeybee Apis mellifera [38] , the parasitic wasp Nasonia vitripennis [39] , Coleoptera: the red flour beetle , Tribolium castaneum [40 , 41] ) . However , whether these genes are involved in the sex determination pathways to form sexually dimorphic horns in T . dichotomus remains elusive . In addition , when and how dsx interacts with the gene regulatory network ( GRN ) for horn formation to drive cellular activities such as cell growth , cell death and cell movement is also unknown . To understand the developmental and genetic mechanisms underlying sexually dimorphic horn formation in T . dichotomus , we first describe a precise time course of the morphogenetic changes of male and female horn primordia during larval-pupal development using time-lapse photography . Next , we examined the function of putative sex determination genes in T . dichotomus using larval RNAi by focusing on orthologs of known D . melanogaster sex determination genes . Moreover , we investigated the initiation timing of the GRN for horn formation by knocking down Tdic-tra at multiple developmental timepoints , and by evaluating the extent of sex transformation phenotypes . Based on these experiments , we concluded that the GRN for horn formation , which is supposed to be modified by dsxM and dsxF functions in males and females , is driven at a very early stage of larval-pupal development before clear morphological changes in horn primordia can be detected . Furthermore , we show that dsxM has different functions in both the prepupal and pupal stages during the formation of appropriate morphologies in pupal and adult horns in males . Based on these findings , we discuss the evolutionary developmental background of sexually dimorphic horn formation in horned beetles .
To identify the developmental timepoint when sexual dimorphism of horns first appears in T . dichotomus , we described morphological changes of head and thoracic horn primordia during the prepupal period . Micro-CT analysis of a head horn primordium at an early stage ( 24 hours after pupal-chamber formation; 24 h APF ) in the prepupa revealed that a head horn primordium was formed in the clypeolabral region during the prepupal stage as investigated in Onthophagus taurus , Onthophagus sagittarius and Onthophagus gazella ( Fig 1A and 1B , S1 Movie ) [3 , 42–44] . In addition , head horn primordia were formed above the clypeus in the clypeolabrum ( Fig 1A , S1 Movie ) . To determine the exact timepoint when protrusion of the primordium is initiated , we established a time-lapse photography system ( Fig 1C , S2 Movie ) . Until recently , developmental staging of T . dichotomus prepupae had been difficult because they form pupation chambers and pupate underground . Using our time-lapse photography system , we found that the head-rocking behavior at the end of pupal-chamber formation can be an unambiguous marker for the initiation of the prepupal stage ( S2 Movie ) . We could minimize the developmental deviation between individuals within 10 hours using this precise developmental marker ( Fig 1C , S2 Movie ) . The average prepupal period was 5 . 5 ± 0 . 19 days ( 131 ± 4 . 7 hours ) in males and 5 . 4 ± 0 . 17 days ( 129 ± 4 . 3 hours ) in females ( Fig 1D ) . Based on this staging paradigm , we manually dissected out horn primordia every 12 h after pupal-chamber formation ( APF ) . We found that sexual dimorphism of horn primordia appeared at 36 h APF ( Fig 1B ) . Therefore , we concluded that the GRN driving the formation of horn sexual dimorphism would be activated before 36 h APF in T . dichotomus . In addition , we also found that apolysis occurring at 36 h APF can be another unambiguous developmental marker . Larval mandibular tendons that tightly connect mandibular muscles and apodemes ( S3 Movie ) [44–46] were completely detached at 36 h APF . This feature , along with the apolysis occurring at every body part including the neighboring ocular region also allowed us to know the timing of the onset of sexual dimorphism in these beetles . In T . dichotomus , the regulatory factors associated with sex-specific splicing of dsx had not been identified . We searched for such regulatory factors focusing on T . dichotomus orthologs of known D . melanogaster sex determination genes ( Sxl , tra , tra2 and ix ) [12 , 22] . First , we investigated whether these genes produce sex-specific splicing variants in T . dichotomus by RT-PCR . All of these genes were expressed in male and female prepupal head and thoracic horns ( Fig 2A light green bars , Fig 2B , S1 Table ) [47 , 48] . Among these genes , sex-specific splicing variants were detected only in Tdic-tra ( Fig 2B ) . Next , to test whether these genes function as sex determination genes , we performed larval RNAi experiments using dsRNA targeting the common regions between sexes ( Fig 2A , black bars , S1 Table , S2 Table ) . In females , morphological changes was observed in the RNAi treatments targeting Tdic-tra and Tdic-ix , whereas no morphological changes were observed in males ( Fig 3A , S2 Fig ) . Such female-specific phenotypes were comparable with the mutant phenotypes of tra , tra2 and ix in D . melanogaster [12 , 22] . Concerning Tdic-tra2 , we could not observe adult phenotypes due to the prepupal lethality of RNAi injection in both sexes . The phenotypes in Tdic-tra RNAi females and Tdic-ix RNAi females were different and the Tdic-ix RNAi phenotype was similar to the effects of the Tdic-dsx RNAi phenotype ( Fig 3A ) . In the Tdic-tra RNAi females , ectopic horn formation was observed in both the head and prothorax ( Fig 3A ) , whereas in Tdic-ix RNAi females , ectopic horns were formed only in the head , and these horns were significantly shorter than the ectopic head horns in Tdic-tra RNAi females ( Fig 3A and 3B ) . Such a difference in morphology was also observed in another sexually dimorphic structure , the intercoxal process of the prosternum ( IPP ) . Male IPPs are generally larger than female IPPs ( Fig 3A ) . Although both the Tdic-tra and Tdic-ix RNAi females have larger IPPs than female controls ( EGFP RNAi ) , Tdic-tra RNAi females have much larger IPPs than Tdic-ix RNAi females ( Fig 3A ) . These results and the intermolecular interactions reported in D . melanogaster led us to predict that Tdic-tra may also regulate the sex-specific splicing of Tdic-dsx in T . dichotomus . To test this , we investigated the splicing patterns of Tdic-dsx in the above RNAi-treated males and females . We designed PCR primer sets to amplify the region including the whole female specific exon , which is spliced out in males ( Fig 2A , light green bars , S1 Table ) . We found that the sex-specific splicing pattern observed in wild type ( Fig 2B ) was switched in Tdic-tra and Tdic-tra2 RNAi treatments , whereas splicing patterns were not changed by Tdic-Sxl and Tdic-ix RNAi treatments ( Fig 3C ) . These data indicate that Tdic-tra and Tdic-tra2 regulate female-specific splicing of Tdic-dsx in T . dichotomus . The switching of sex-specific splicing of Tdic-dsx and the resultant morphological sex transformation in Tdic-tra RNAi females implies that the sex-specific splicing regulation of Tdic-dsx by Tdic-tra is also conserved in T . dichotomus as in other holometabolous insects investigated so far [33–41] . Taken together , we concluded that Tdic-tra functions as a sex determination gene during horn formation in T . dichotomus . As sexual dimorphism of horn primordia first appeared at 36 h APF ( Fig 1B ) , we speculated that the onset of the developmental program for sexually dimorphic horn formation was initiated before 36 h APF . To estimate this timepoint more accurately , we performed Tdic-tra RNAi in females at multiple developmental timepoints during pupal chamber formation periods and prepupal periods , and evaluated the extent of sexual transformation in horns . If the timing of Tdic-tra RNAi in females is early enough , the ectopic Tdic-dsxM would be expressed in female horn primordium from the onset of developmental program for sexual dimorphism formation , and full sexual transformation can be achieved . On the other hand , the later the timing of Tdic-tra RNAi treatment becomes , the more the initial phases of the male-specific horn formation program driven by ectopically expressed Tdic-dsxM are trimmed , and at the same time repressed by normally expressed dsxF . Therefore , by determining the latest RNAi injection timing when a full sexual transformation phenotype is observed , we can estimate the onset of the sexually dimorphic horn formation program mediated by the ectopic Tdic-dsxM expression . In this experiment , fully matured female last instar larvae and unstaged female prepupae were injected with EGFP or Tdic-tra dsRNA ( S3 Table ) . We performed time-lapse photography until pupation using these larvae , and retrospectively estimated the exact timing of injection before pupation . In addition , we estimated how many hours after pupal chamber formation ( APF ) each injection had been performed by subtracting the duration between each injection timepoint and pupation timepoint from the mean duration of the prepupal period in Tdic-tra RNAi females ( 127 hours ) ( Fig 1D ) . Our data indicated that Tdic-tra RNAi females treated earlier than the timepoint of -7 h APF formed fully developed male horns ( Fig 4A ( i ) , magenta dot ) . In contrast , Tdic-tra RNAi females treated later than the timepoint at -3 h APF showed either no morphological changes or only modest sex transformation of head or horn , if any ( Fig 4A ( ii ) and ( iii ) , green dot ) . These data suggested that the estimated timepoint of the onset of the developmental program for sexually dimorphic horn formation is around -7 h APF . Since this timepoint was estimated by means of RNAi , we speculated that there should be time lag between the timing of injection and the timing of the decrease in functional protein levels followed by the mRNA degradation . Thus , we also quantified expression dynamics of Tdic-tra mRNA after RNAi treatment by qRT-PCR . It was technically impossible to monitor the expression dynamics of Tdic-dsx at -7 h APF , which is estimated to be before pupal chamber formation ( Fig 1D , Fig 4A ) . Then , we monitored the expression dynamics as early as possible ( dsRNA injection at 24 h APF ) , instead . The expression levels of Tdic-tra and Tdic-dsxF were quantified every 12 hours up to 36 hours after injection ( Fig 2A , green bars , Fig 4B ) . As a result , the expression levels of Tdic-tra and Tdic-dsxF were decreased to less than half of that of the negative control at 36 hours after injection ( Fig 4B ) . Since mRNA started to be degraded 36 hours after RNAi treatments ( Fig 4B ) , we estimated that this timepoint corresponded to 29 h APF ( −7 plus 36 ) ( Fig 4C ) . Taking morphological data into account , this timepoint corresponded to 7 hours before the initial sexual dimorphism appears in horn primordia ( Fig 1B , Fig 4C ) . The expression level of Dsx protein is frequently upregulated region-specifically during development of sexually dimorphic traits in insects ( e . g . Sex comb formation in Drosophila , mimetic wing morph formation in Papilio , and wing pheromone gland formation in Bicyclus ) , presumably to facilitate sexual dimorphism formation [18–21] . To test whether Tdic-Dsx protein also exhibits region-specific upregulation in the horn primordium , we raised anti-Tdic-Dsx polyclonal antibodies , and performed immunohistochemistry at the onset of sexually dimorphic horn formation ( 36 h APF ) . As a result , Tdic-Dsx protein showed higher expression in the head primordial epidermis than in the surrounding head epidermis and was mainly localized in nuclei ( Fig 5D–5F and 5D’–5F’ , S3 Fig ) . In accordance with this result , mRNA expression level of Tdic-dsxM were also higher in the head horn primordial epidermis than in the surrounding head epidermis ( Fig 5G and 5H ) . In contrast , Tdic-dsxF did not show significantly higher expression in the horn primordium at this stage ( Fig 5G and 5H ) . Therefore , we concluded that expression level of Dsx protein is upregulated region-specifically to form sexually dimorphic horns during development . Interestingly , we found that Tdic-Dsx showed higher expression in the head horn primordial epidermis even before the onset of sexually dimorphic horn formation ( 12 h APF ) ( Fig 5A–5C and 5A’–5C’ , S3 Fig ) , but the localization of expression was cytoplasmic . This finding suggests that the region-specific expression of Tdic-Dsx has been already initiated before it activates downstream genes , but its transcription factor activity is triggered only after it is translocated to nuclei . T . dichotomus adult males have exaggerated horns at the head and prothoracic regions whereas adult females only have three small protrusions at the clypeolabral region ( Fig 6A , magenta arrowhead ) . We focused on whether the distal tips of the male head horn and three female head protrusions were formed in the same region or whether they originated from distinct regions in the head ( Fig 6A , S1 Fig ) . During prepupal stages , a head horn primordium in both males and females seemed to be formed in the same region ( the almost entire clypeolabral region in the head ) ( Fig 1 , S1 Fig ) . However , due to the lack of clear morphological landmarks indicating the formation region of male head horns and female head protrusions , within the head , the formation regions of these traits remained elusive . Unexpectedly , however , we obtained an intermediate sexual transformation phenotype of Tdic-tra RNAi to solve this problem . When we injected Tdic-tra dsRNA in small amounts ( 2 . 5 μg ) , we obtained several adults possessing both the three small protrusions similar to those in females , and a small ectopic anterior protrusion , seemingly analogous to a male head horn ( Fig 6B and 6C ) . This result suggests that at least the anterior region of a male head horn is not formed from the same region as female head protrusions . Next , we asked whether Tdic-dsx regulates the entire cellular activity during sexually dimorphic horn formation . We especially focused on the two distinct developmental processes , “horn growth” before pupation , and “horn remodeling” after pupation ( see details in the Introduction section ) . The morphological changes before and after pupation are qualitatively comparable between males and females , but the extent of the morphological changes during development is different between them . During the prepupal stage , males form longer head and thoracic pupal horns , whereas females only form smaller pupal head horns . During the pupal stage , both male head and thoracic horns become slimmer , whereas a substantial portion of the female pupal head horn disappears to form small three protrusions ( S1 Fig ) . We tested whether Tdic-dsx regulates the above sexually dimorphic morphogenetic processes by injecting dsRNA targeting Tdic-dsx into male larvae at several developmental stages ( S3 Table ) . As previously reported , male head horns became shorter , and thoracic horns were not formed in adults when Tdic-dsx dsRNA was injected at sufficiently early timepoints [8] . In these conditions , the sizes of the pupal head horns and the pupal thoracic horns were also smaller proportionally to that of shortened adult horns ( -85 h APF ) ( Fig 7B , S4 Fig ) . On the other hand , males treated with Tdic-dsx dsRNA at later stages ( 13 h APF ) formed thickened adult thoracic horns similar to a pupal thoracic horn before remodeling ( Fig 7C ) and the head horn was comparable with the wild type head horn ( S4 Fig ) . These data indicate that DsxM is required for horn remodeling only in the thorax , and is dispensable in the head horn remodeling .
As mentioned in the Introduction section , the function of regulatory factors that are supposed to bind directly to dsx transcripts ( Tra , Tra2 ) or Dsx protein ( Ix ) are conserved in many holometabolous insects . Our loss-of-function analysis data suggest that the genetic regulatory mechanisms of sex determination in T . dichotomus follow this framework . The conserved function of Tdic-Tra and Tdic-Tra2 as splicing factors targeting Tdic-dsx seems to be conserved ( Fig 3C ) . In addition , the biological functions of these genes are at least similar to those in other beetles because the RNAi phenotypes of tra and tra2 orthologs in T . dichotomus ( i . e . the viable masculinized phenotype in Tdic-tra RNAi females , and the lethal phenotype in Tdic-tra2 RNAi females ) were comparable with those reported in other beetles ( T . castaneum; [40 , 41] , the stag beetle Cyclommatus metallifer; [49] ) . In D . melanogaster , Ix directly binds to DsxF in females but not to DsxM in males , and functions as a co-activator to facilitate transcription activity of target genes of DsxF [32] . The intersexual phenotype observed only in Tdic-ix RNAi-treated females ( Fig 3A ) implies that Tdic-Ix might also interact with Tdic-DsxF to regulate female-specific sex differentiation as in D . melanogaster . A female-specific intersexual transformation phenotype reported in the stag beetle C . metallifer [49] suggests that the female-specific function of ix may be conserved among Polyphaga as well . In contrast , our RNAi experiments implied that Tdic-Sxl did not regulate the sex-specific alternative splicing of Tdic-tra in females ( Fig 3 ) , which is a direct regulatory target of Sxl in D . melanogaster females . Because the splicing of tra orthologs are not regulated by Sxl orthologs even in other Dipteran species [50 , 51] , the sex-specific alternative splicing of tra may be regulated by unknown species-specific factors other than Tdic-Sxl . To summarize the regulatory mechanisms of T . dichotomus sex determination discussed above , functional Tdic-Tra is first expressed female-specifically through species-specific unknown mechanisms . Then , as in many other holometabolous insects , a Tdic-Tra/Tdic-Tra2 heterodimer is formed only in females and produces Tdic-dsxF mRNA and Tdic-DsxF to promote female differentiation by interacting with Tdic-Ix ( Fig 8 ) . On the other hand , in males , functional Tdic-Tra is not expressed , and default splicing of Tdic-dsx would result in Tdic-dsxM mRNA and Tdic-DsxM production to promote male differentiation ( Fig 8 ) . A previous study revealed that Tdic-dsx RNAi in both males and females does not result in a total loss of horns , but results in intermediate-lengthened head horn formation and loss of a thoracic horn ( Fig 3 ) [8] . The head horn phenotype suggests that there exists a GRN for head horn formation that is operated independently of Tdic-dsx . Tdic-dsxM seems to enhance this GRN whereas Tdic-dsxF seems to suppress it . On the other hand , the thoracic horn phenotypes suggest that thoracic horn formation seems to be totally dependent on Tdic-dsxM function , and that action of Tdic-dsxM on horn formation might be different between head and thoracic horns during prepupa and/or pupa . However , in the course of our Tdic-dsx RNAi experiments , we noticed that thoraces of severely intersexually transformed individuals are always slightly bulged at the thoracic horn formation region ( Fig 3A ) . This implies that Tdic-dsxF expression in the female thorax still has suppressive activity against the GRN for thoracic horn formation . If this is the case , the regulatory relation between Tdic-dsx and GRN for horn formation has partial similarity between the head and the thorax . The phenotypic difference previously described in head and thoracic horns would be due to difference in length of horn formed in each body region . In the sections below , we mainly discuss the GRN for horn formation based on our head horn data in which interpretation of phenotypes is less ambiguous . Our knockdown experiment of Tdic-tra revealed that the onset of Tdic-dsx modulating GRN for head horn formation is as early as 29 h APF ( Fig 4 ) . This is approximately 7 hour before the appearance of sexual dimorphism of head horn primordium ( Fig 1B ) . Our qRT-PCR and immunohistochemistry analysis focusing on Tdic-Dsx were consistent with this estimation: transcription and translation of dsx was initiated before 0 h APF , but nuclear translocation of Dsx was initiated betweeen 12 h and 36 h APF ( Fig 5 , S3 Fig , S5 Fig ) . Therefore , Tdic-dsx seems to modulate head horn primordium formation just before the onset of tissue growth , and following tissue morphogenesis during prepupal stages . Several genes involved in the GRN for horn formation have been identified in horned beetle species , primarily by focusing on Drosophila appendage patterning genes [52–54] . These studies propose an attractive evolutionary model in which large portions of the GRN for proximodistal appendage patterning were recruited to acquire beetle horns in the head and thoracic regions . Still , the overall framework of GRN for horn formation including the key regulatory gene sets involved , spatiotemporal expression dynamics , and regulatory relation among those genes , has not been unveiled so far . Therefore , investigating regulatory relations between appendage patterning genes and other head/thoracic patterning genes expressed in horn primordia focusing on this developmental timepoint will lead to verification of this model in the future studies . In addition , because candidates of the dsx target genes involved in GRN for horn formation were identified recently in a horned beetle , O . taurus [55] , functional analyses focusing on orthologs of those genes expressed in Trypoxylus horn primordia at this timepoint will also lead to understanding of conserved and divergent aspects of sexually dimorphic horn formation in horned beetles . Region-specific upregulation of dsx is another essential feature to drive GRN for sexual dimorphism formation in insects [17–21] . Intense studies focusing on regulatory mechanisms of sex comb formation in D . melanogaster revealed that such region-specific upregulation of dsx is mediated by a positive feedback loop between dsx and a Hox gene , Sex combs reduced [18] . Therefore , region-specific upregulation of Tdic-dsx observed in a head horn primordium might reflect an analogous positive feedback loop mechanism ( Fig 5 ) . Future research focusing on the patterning mechanisms of beetle horns and its interaction with dsx will be an important issue to discuss conserved genetic framework for sexual dimorphism formation . Moreover , we found nuclear translocation of Tdic-Dsx at the onset of male head horn formation ( Fig 5 , S3 Fig ) . As far as we know , such a mode of regulation against Dsx during development is not reported in other insects . Understanding the molecular mechanisms underlying this phenomenon will be another direction to take in the future studies . As previously reported , Tdic-dsx RNAi from early stages resulted in short head horn formation in males and females ( Fig 3A ) [8] . These data indicate that during the prepupal stage Tdic-dsxM promotes the growth of head and thoracic horn primordium , whereas Tdic-dsxF suppresses growth of head and thoracic horn primordium ( Fig 8 , prepupal stage ) [8] . On the other hand , RNAi treatments in later stages revealed distinct functions of Tdic-dsx in sexually dimorphic horn formation . During pupal-adult development , Tdic-dsxM regulates remodeling of a thoracic horn primordium from a rounded shape to a slender hooked shape ( Fig 7 , Fig 8 , pupal stage ) , but is dispensable for head horn remodeling ( S4 Fig ) . These results indicate that Tdic-dsxM and Tdic-dsxF regulate different aspects of morphogenesis at both prepupal and pupal stages . Furthermore , to what extent Tdic-dsx is required during larval-pupal development is also finely regulated during male head horn formation , male thoracic horn formation , female head protrusion formation , and female flat prothorax formation . Importantly , in contrast to the previous study that suggested that functions of Tdic-dsxM and Tdic-dsxF in T . dichotomus head and thoracic horn formation are to some extent analogous as described in the previous section ( that is , in both of the head and the thoracic horn formation , Tdic-dsxM functions as a positive regulator whereas Tdic-dsxF functions as a negative regulator ) [8] , our experiments clarified that both spatial cues ( i . e . different developmental contexts in head and prothorax ) and temporal cues ( i . e . different developmental contexts in prepupa and pupa ) modulate the actions of Tdic-dsxM and Tdic-dsxF to drive appropriate morphogenetic activity in each horn primordium at each developmental stage . Since spatiotemporal modularity of gene function is often coded in modular regulatory elements in the genome ( reviewed in [56]; e . g . [57–60] ) , the regulatory elements of dsx that integrates the above distinct spatiotemporal cues could be coded at the Tdic-dsx locus in the genome . Identification of such regulatory elements will be an important issue to elucidate evolution of sexually dimorphic horn formation in the future studies . A notable feature of Tdic-dsx during sexually dimorphic development is that the onset of action during horn development seems to be categorized earlier ( i . e . during the prepupal period ) than that of many other sexually dimorphic adult traits in holometabolous insects reported so far ( i . e . the pupal period ) . In D . melanogaster sex comb formation , Papilio mimetic wing morph formation , and Bicyclus wing pheromone gland formation , tissue-specific high expression of dsx orthologs are detected after pupation [18–21] . Such a difference in dsx’s onset of action seems to be due to the requirement of drastic growth during sexually dimorphic structure formation . In either case mentioned above , the finally formed structure accompanies little to no tissue-level drastic growth during development . On the other hand , as in T . dichotomus horn formation , D . melanogaster genital organ formation , and C . metallifer stag beetle mandible formation , whose sexually dimorphic morphogenesis is regulated by dsx , accompanies drastic sexually dimorphic growth during larval-pupal development [17 , 61–63] , and its onset of action is as early as the prepupal stage . Association between dsx’s earlier onset of action and requirement of drastic growth during sexual dimorphism formation suggest that recruitment of the dsx function in the earlier stage may be one of the prerequisites to form structurally drastically different sexual dimorphism in insects . The intermediate head horn formation via Tdic-tra RNAi in females ( Fig 6A–6C ) revealed that the anterior region of male head horn and female small protrusions are formed from different regions in the clypeolabrum . Such a distinct developmental origin of head horn within the clypeolabrum is also clearly demonstrated in Onthophagus species of dung beetles [3] , in which the dsx orthologs regulate the formation of distinct sexually dimorphic horns in either the anterior or the posterior region of the head . Analogy of multiple horn formation regions within the clypeolabrum in distinct horned beetle species implies that clypeolabrum is a hotspot of morphological innovations in horned beetles . A comprehensive understanding of the GRN for horn formation in both Trypoxylus and Onthophagus and comparative developmental studies in the future will lead to understanding of molecular mechanisms underlying the evolutionary origin and evolvability of exaggerated horns in beetles . Here we described the accurate developmental time course of horn primordial morphogenesis during the prepupal stage in the horned beetle T . dichotomus using a time-lapse photography system . In addition , we functionally characterized both Tdic-tra and Tdic-tra2 , genes that regulate the sex-specific splicing of Tdic-dsx . By manipulating expression levels of Tdic-tra and Tdic-dsx during different developmental time points , and by quantifying the extent of sex transformation , we revealed the following three crucial features of Tdic-dsx function during the development of sexually dimorphic horn formation: ( 1 ) Tdic-dsx modulates the GRN for horn formation as early as 29 h APF , a timepoint which corresponds to 7 h before sexual dimorphisms of horn primordia first appears; ( 2 ) Tdic-dsx regulates different aspects of the tissue growth , tissue death and tissue movement of horn primordia depending on both spatial ( head/prothorax ) and temporal ( prepupa/pupa ) contexts; ( 3 ) Tdic-dsxM and Tdic-dsxF promotes the formation of outgrowth structure in distinct regions within the clypeolabrum . These findings inform our understanding of the patterning mechanisms at play during T . dichotomus horn formation , as well as provide information regarding regulatory shifts in these mechanisms during the evolution of sexually dimorphic traits in horned beetles . The present study provides a good starting point to elucidate such issues .
We purchased T . dichotomus larvae from Loiinne ( Japan ) , and Urakiso Tennen Kabuto no Sato ( Japan ) . The last instar larvae were sexed as described previously [8] , individually fed on humus in plastic containers , and kept at 10°C until use . Larvae were moved to room temperature at least 10 days , and reared at 28°C . A male head tissue at 24 hours after pupal-chamber formation ( 24 h APF ) was fixed in Carnoy solution at room temperature overnight , washed in 70% ethanol and stored in 70% ethanol . The sample was rehydrated through a graded ethanol series , and stained with 25% Lugol solution [64–66] for 5 days . The stained sample was scanned using an X-ray micro-CT device ( ScanXmate-E090S105 , Comscantechno Co . , Ltd . , Japan ) at a tube voltage peak of 60 kVp and a tube current of 100 μA . The sample was rotated 360 degrees in steps of 0 . 24 degrees , generating 1500 projection images of 992 × 992 pixels . The micro-CT data were reconstructed at an isotropic resolution of 13 . 3 × 13 . 3 × 13 . 3 μm , and converted into an 8-bit tiff image dataset using coneCTexpress software ( Comscantechno Co . , Ltd . , Japan ) . Three-dimensional tomographic images were obtained using the OsiriX MD software ( version 9 . 0 , Pixmeo , SARL , Switzerland ) and Imaris software ( version 9 . 1 , Carl Zeiss Microscopy Co . , Ltd . , Japan ) . Supplemental videos were edited using Adobe Premiere Pro CC ( Adobe Systems Co . , Ltd . , Japan ) . To monitor the precise time course of the morphogenetic changes of male and female horn primordia during larval-pupal development , time-lapse photography was performed at 28°C every 30 minutes until they had developed into adults using a CMOS camera ( VCC-HD3300 , SANYO , Co . , Japan ) . We found that the head-rocking behavior observed at the end of pupal-chamber formation can be an unambiguous developmental marker for initiation of the prepupal stage ( 0 h APF ) . Using this marker , sampling of horn primordia for staging was performed every 12 hours until they had developed into pupae ( 120 h APF ) using time-lapse photography . Head and thoracic horn primordia were dissected out from prepupae in 0 . 75% sodium chloride , and fixed for 90 minutes at room temperature with 4% paraformaldehyde in phosphate buffered saline ( PBS ) . After being washed twice in PBS , photographic images were obtained with a digital microscope ( VHX-900 , KEYENCE , Co . , Japan ) . We searched for orthologs of the sex determination genes in the T . dichotomus RNA-seq database ( PRJDB6456 ) using full-length complementary DNA ( cDNA ) sequences of D . melanogaster genes ( Sxl , tra , tra2 , ix ) as query sequences using the tblastn program ( https://blast . ncbi . nlm . nih . gov/Blast . cgi ) . We evaluated orthology of those genes by performing reciprocal tblastx searches against the D . melanogaster cDNA database ( r6 . 06 ) using the identified T . dichotomos genes as queries . We deposited cloned partial cDNA sequences of T . dichotomus genes at the DNA Data Bank of Japan ( DDBJ ) /European Molecular Biology Laboratory ( EMBL ) /GenBank . The accession number for Tdic-Sxl is LC385009 , for Tdic-tra are LC385010—LC385012 , for Tdic-tra2 are LC385013—LC385018 , and for Tdic-ix is LC385019 . Total RNA was extracted from each of head or thoracic horn primordia in wild type and RNAi-treated individuals ( EGFP , Sxl , tra , tra2 , ix , dsx and dsxF ) using TRI Reagent ( Molecular Research Center , Inc . , USA ) according to the manufacturer’s instructions . First-stranded cDNA was synthesized with the SuperScript III Reverse Transcriptase ( Life Technologies Japan Ltd . , Japan ) using 1 μg total RNA as a template . Primer sets for cloning and double stranded RNA ( dsRNA ) synthesis were designed based on the cDNA sequences identified above ( S1 Table ) . PCR was performed using Ex Taq DNA polymerase ( Takara Bio Inc . , Japan ) according to the manufacturer’s protocol . Amplified PCR products were purified using MagExtractor ( TOYOBO , Co . , Ltd . , Japan ) , and subcloned into the pCR4-TOPO vector using the TOPO TA cloning Kit ( Life Technologies Japan Ltd . , Japan ) . Sequences of the inserts were determined by a DNA sequencing service at FASMAC Co . Ltd . , Japan . Primer sets for qRT-PCR were designed using Primer3Plus program ( http://primer3plus . com/cgi-bin/dev/primer3plus . cgi ) ( S1 Table ) . qRT-PCR was performed using the cDNA libraries synthesized above and THUNDERBIRD SYBR qPCR Mix ( TOYOBO , Co . , Ltd . , Japan ) according to the manufacturer’s instructions . The relative quantification in gene expression was determined using the 2-ΔΔCt method [67] . Primer pairs for dsRNA synthesis were designed within the common regions shared by male and female isoforms ( Fig 2A , S1 Table ) . Partial sequences of the target sequences were amplified by PCR using primers flanked with the T7 promoter sequence in the 5’-ends . DsRNAs were synthesized using AmpliScribe T7-Flash Transcription Kit ( Epicentre Technologies , Corp . , USA ) . The purified PCR products were used as templates . Injections of dsRNA were performed as described previously [8] . DsRNA was injected into each late last instar larva or prepupa under the conditions described in S2 Table and S3 Table . Enhanced green fluorescent protein ( EGFP ) dsRNA was injected as a negative control . Horn length was measured by extracting the contour of a horn in the lateral view using the SegmentMeasure plugin for ImageJ 64 developed by Hosei Wada . We defined a body length as the length between the anterior tip of the clypeus in the head to the posterior most region of the abdomen , and was directly measured using a digital caliper ( DN-100 , Niigata seiki , Co . , Ltd . , Japan ) . Relative horn length was calculated by dividing the horn length by the body size in each RNAi treated individual , and standardized by dividing by the mean horn length of the EGFP RNAi-treated males . Differences of medians between treatment groups were evaluated with Brunner-Munzel test using the R package "lawstat" ( ver . 3 . 2 ) . These p-values were adjusted by the Bonferroni correction . A DNA fragment encoding the N-terminal region of Tdic-Dsx ( 1–139 a . a . ) was amplified with the primer flanked with an NcoI restriction site 5′-CCATGGCCGACTCGCAAGAGTACGAAGCCA-3′ , and the primer flanked with BamHI restriction site 5′-GGATCCTTAGTTATTACCAACGGTTTCCCG-3′ ( each restriction site is underlined ) , and inserted into each of two vectors , pET-32b and pET-41b ( Merck KGaA . , Germany ) , in order to express Trx- and GST-fused recombinant proteins ( Trx–Tdic-Dsx and GST–Tdic-Dsx ) in Escherichia coli , BL21 ( DE3 ) ( New England Biolabs , Inc . , Japan ) . Trx–Tdic-Dsx and GST–Tdic-Dsx were recovered from SDS polyacrylamide gel after electrophoresis , and stored at -20°C until use . Guinea pig polyclonal antibodies were raised against the GST–Tdic-Dsx , and affinity-purified with a Hitrap NHS-activated HP column ( GE Healthcare UK Ltd . , UK ) coupled with the above Trx–Tdic-Dsx fusion protein . After dissection , head horn primordia fixed with 4% PFA was embedded in SCEM compound ( Leica Microsystems , GmbH . , Japan ) and stored at -80°C until use . The frozen block was sectioned transversely at 10 μm using Leica CM1950 ( Leica Microsystems , GmbH . , Japan ) , and washed with PBS after brief drying . Guinea pig anti-Tdic-Dsx antibody was used as the primary antibody at a 1:500 dilution . Biotin-labeled goat anti-guinea pig IgG antibody ( Jackson ImmunoResearch Laboratories , Inc . , USA ) was used as the second antibody at a 1:500 dilution . For fluorescence staining , streptavidin-HRP and Cy3-conjugated tyramide ( PerkinElmer Japan Co . , Ltd . , Japan ) were used at dilutions of 1:5000 , 1:500 , respectively . Laser scanning confocal microscopy ( Olympus FV-1000 ) was used to visualize immunostained frozen tissue sections . Tiled array images were obtained using a confocal microscopes equipped with a motorized stage ( Olympus 3D mosaic imaging system ) .
|
Beetles in the family Scarabaeidae have various types of horns on their heads and thoraces , and the shape , size , number , and location of these horns are highly diversified within the group . In addition , many scarab beetle horns are sexually dimorphic . The acquisition of these evolutionarily novel horns , and the mechanisms for the diversification of these structures is an interesting question . To address this question , we focused on the rhinoceros beetle Tripoxylus dichotomus . Here we identified the exact developmental timepoints during which the morphological sexual dimorphism of horn primordia appears , estimated the onset of the developmental program for sexually dimorphic horn formation driven by doublesex , and revealed that doublesex regulates different aspects of cell activities during horn formation depending on particular spatiotemporal developmental contexts . Our study provides insights into regulatory shifts in these mechanisms during the evolution of sexually dimorphic traits in horned beetles .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"invertebrates",
"rna",
"interference",
"gene",
"regulation",
"animals",
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"experimental",
"organism",
"systems",
"epigenetics",
"sexual",
"dimorphism",
"morphogenesis",
"drosophila",
"research",
"and",
"analysis",
"methods",
"genetic",
"interference",
"animal",
"studies",
"gene",
"expression",
"beetles",
"insects",
"sex",
"determination",
"arthropoda",
"biochemistry",
"rna",
"eukaryota",
"nucleic",
"acids",
"sexual",
"differentiation",
"genetics",
"biology",
"and",
"life",
"sciences",
"evolutionary",
"biology",
"organisms",
"evolutionary",
"developmental",
"biology"
] |
2019
|
Precise staging of beetle horn formation in Trypoxylus dichotomus reveals the pleiotropic roles of doublesex depending on the spatiotemporal developmental contexts
|
Transcriptional enhancers play critical roles in regulation of gene expression , but their identification in the eukaryotic genome has been challenging . Recently , it was shown that enhancers in the mammalian genome are associated with characteristic histone modification patterns , which have been increasingly exploited for enhancer identification . However , only a limited number of cell types or chromatin marks have previously been investigated for this purpose , leaving the question unanswered whether there exists an optimal set of histone modifications for enhancer prediction in different cell types . Here , we address this issue by exploring genome-wide profiles of 24 histone modifications in two distinct human cell types , embryonic stem cells and lung fibroblasts . We developed a Random-Forest based algorithm , RFECS ( Random Forest based Enhancer identification from Chromatin States ) to integrate histone modification profiles for identification of enhancers , and used it to identify enhancers in a number of cell-types . We show that RFECS not only leads to more accurate and precise prediction of enhancers than previous methods , but also helps identify the most informative and robust set of three chromatin marks for enhancer prediction .
Enhancers are distal regulatory elements with key roles in the regulation of gene expression . In higher eukaryotes , a diverse repertoire of transcription factors bind to enhancers to orchestrate critical cellular events including differentiation [1] , [2] , maintenance of cell-identity [3] , [4] and response to stimuli [5]–[7] . While enhancers have long been recognized for their regulatory importance , the fact that they lack common sequence features and often reside far away from their target genes has made them difficult to identify . Computational techniques relying on transcription factor motif clustering or comparative analyses have had some success in identifying enhancers , but these predictions are neither comprehensive nor tissue-specific [8]–[13] . Recently , several high-throughput experimental approaches have been developed to identify enhancers in an unbiased , genome-wide manner . The first is mapping the binding sites of specific transcription factors by ChIP-seq [14] . Because this approach requires the knowledge of a subset of transcription factors ( TFs ) that are not only expressed but also occupy all active enhancer regions in the cell-type of interest , identification of all enhancers using this approach is not a trivial task . The second approach involves mapping the binding sites of transcriptional co-activators such as p300 and CBP [4] , [5] , [15] , which are recruited by sequence-specific transcription factors to a large number of enhancers [6] , [16] , [17] . Since not all enhancers are marked by a given set of co-activators [18] , [19] , and ChIP-grade antibodies against these proteins may not always be available , systematic identification of enhancers by mapping the locations of co-activators is not generally feasible . A third approach relies on identifying open chromatin with techniques such as DNase I hypersensitivity mapping [20] . However , since open chromatin regions can correspond to not only enhancers , but also silencers/repressors , insulators , promoters [21] , [22] or other functionally unknown sequences occupied by nuclear proteins , this approach lacks specificity in enhancer identification . Finally , a fourth approach interrogates covalent modifications of histones [5] , [23]–[26] as it was observed that certain histone modifications form a consistent signature of enhancers . It is on this approach that the present work is focused . Previously , we and others observed that distinct chromatin modification patterns were associated with transcriptional enhancers [5] , [22] , [27] . Specifically , active promoters are marked by trimethylation of Lys4 of histone H3 ( H3K4me3 ) , whereas enhancers are marked by monomethylation , but not trimethylation , of H3K4 ( H3K4me1 ) . This chromatin signature has been used to develop a profile-based method for enhancer discovery [5] . Both unsupervised [25] , [28] and supervised learning approaches have also been employed to exploit chromatin modification-based differences to identify enhancers . The supervised machine learning techniques include HMM [8] , [23] , neural networks [24] and genetic algorithm-optimized SVM [26] based approaches , and have proved to be improvements over the profile-based method . While these methods have led to identification of a great number of enhancers in the human and mouse genomes [3] , [25] , [29] , the current computational techniques have thus far been limited by the small number of the training set samples and limited number of chromatin modifications examined . Thus , it is possible that these approaches may not fully capture the entire range of chromatin modification patterns at enhancer elements . With the discovery of ever more histone modifications , it is likely that additional chromatin modifications may distinguish enhancers from other functional elements in the genome . This additional data should in principle allow us to answer the key question: what is the optimal set of modifications required for enhancer prediction ? Some researchers have tried to tackle this issue by using algorithms such as simulated annealing [23] or genetic-algorithm optimization [26] . We sought to develop a method in which the selection of the optimal set is automatically built into the training-process and is easily adapted to a large number of features . As part of the NIH Epigenome Roadmap project , we have generated genome-wide profiles for 24 chromatin modifications and DNase-I hypersensitivity sites in 2 distinct cell types- human embryonic stem cell ( H1 ) and a primary lung fibroblast cell line ( IMR90 ) [30] . Additionally , we have experimentally determined a large number of promoter-distal p300 binding sites in each cell type , providing a rich training set for development of accurate and robust enhancer prediction algorithms . We now describe a random-forest [31] based method for integrative analysis of diverse histone modifications to predict enhancers . We show that this new algorithm outperforms the existing methods and leads to the automatic discovery of an optimal set of chromatin modifications for enhancer predictions .
Random forests have recently become a popular machine learning technique in biology [32] due to their ability to run efficiently on large datasets without over-fitting , and their inherently non-parametric structure . Since random forests use a single variable at a time , they can give an automatic measure of feature importance [33] . Hence , we developed an algorithm based on this random forest technique for the purpose of enhancer prediction . Conventional random forests utilize a single scalar value associated with each feature at each node of the tree . In order to train a random-forest for enhancer prediction we wanted to use histone modification profiles at p300 binding sites . Because the spatial organization of histone modifications along a linear chromosome can be as informative as their actual levels , they are better represented as vectors of binned reads . Inspired by recent modifications to the random-forest approach such as discriminant random forests [34] or oblique random forests [35] that utilize a linear classifier at each node , we developed a new vector-based random forest algorithm RFECS or Random Forest for Enhancer Identification using Chromatin States ( see Methods ) . Genome-wide distal p300 binding sites were found using ChIP-seq in H1 and IMR90 cell-lines . We selected p300 binding sites overlapping DNase-I hypersensitive sites and distal to annotated TSS as active p300 binding sites representative of enhancers . We found 5899 such p300 binding sites in H1 and 25109 such sites in IMR90 ( Table S1 , S2 ) , and observed several distinct and diverse chromatin states using an unsupervised clustering technique , ChromaSig ( fig . 1A , B ) . All clusters showed enrichment of H3K4me1 and depletion of H3K4me3 as previously observed [5] . However , different clusters were characterized by varying levels of histone acetylation , H3K4me2 or H3K27me3 . Clusters with presence or absence of H3K36me3 may represent genic and intergenic enhancers respectively . In order to ensure we represented all these different chromatin states at active p300 binding sites , we selected a relatively large number of these sites ( >5000 ) for training as compared to previous methods . To train the forest , active and distal p300-binding sites ( BS ) were selected as representative of the enhancer class . As non-enhancer classes , we considered annotated transcription start sites ( TSS ) that overlap DNase-I , and random 100 bp bins that are distal to known p300 or TSS ( see Methods ) . The confidence of each enhancer prediction is given by the percentage of trees that predict this site to be an enhancer . In general , a genomic region is predicted as an enhancer if it has a background cutoff greater than 0 . 5 ( >50% trees vote in it's favor ) . At higher cutoffs , confidence of prediction is higher , but fewer enhancers are predicted . We used Receiver Operating Characteristic ( ROC ) curves to determine optimal parameters for our classification algorithm [36] . In the case of enhancer predictions , we can only obtain an approximate measure of specificity since we can never be certain that the randomly selected elements of the non-p300 class are all true negatives . Hence , in addition to the ROC curves generated using 5-fold cross-validation , we also verified parameter selection by comparing the percentage of predicted enhancers at each cutoff that overlap markers of active enhancers ( validation rate ) or TSS ( misclassification rate ) . The markers of active enhancers include distal DNase-I hypersensitivity sites ( HS ) , p300 binding sites ( excluding those used in training ) , occupancy by CBP or sequence-specific transcription factors known to act at embryonic stem cell enhancers such as NANOG , OCT4 and SOX2 . In the case of Random forests , the main parameter to be determined is the number of trees . Since the non-enhancer class is assumed to be several times enriched compared to the enhancer class in the genome , we select a greater number of non-p300 training sites as compared to p300 sites and this proportion is also adjusted using the above-described methods . Previous algorithms [23] as well as empirical observations showed a width of −1 kb to +1 kb around the p300 binding site as optimal but we further verified this selection by cross-validation in the H1 cell-type ( fig . S1A ) . The difference in cross-validation curves using a width of 0 . 5 kb or 1 kb is not obvious on the cross-validation curve while a width of 1 . 5 kb clearly shows a sharp drop in the area under the ROC curve ( fig . S1A ) . When we further made enhancer predictions using all three widths ( fig . S1B , C ) , it can be seen that a width of 1 kb on either side shows best validation and misclassification rates as compared to 0 . 5 or 1 . 5 kb widths . To determine the optimal number of trees for the random-forest , we examined the area under the ROC curve in H1 and IMR90 and found both to be stable beyond 45 trees ( fig . 2A , B ) . In order to verify this further , we made enhancer predictions using various number of trees such as 45 , 65 and 85 and compared the validation and misclassification rates ( fig . S2A–D ) . While H1 appeared to show no change at all ( fig . S2A , , C ) IMR90 showed a slight improvement from 45 to 65 trees ( fig . S2B , D ) . In the end , we selected 65 trees for training the random forest as it appeared to be optimal for both cases . The training-set ratio of p300 to non-p300 was set at 1∶7 since the ROC curve did not appear to change much beyond this ratio . ( fig . S2E , F ) In order to estimate the accuracy of the enhancer prediction by RFECS , we applied this algorithm to chromatin profiles of 24 marks obtained in H1 and IMR90 . We then calculated the validation rate as the percentage of predicted enhancers overlapping with DNase-I hypersensitivity sites and binding sites of p300 and a few sequence specific transcription factors known to function in each cell type ( true positive markers ) . We also computed the misclassification rate as the percentage of predicted enhancers overlapping with known promoters . These overlaps were computed using a window of −2 . 5 to +2 . 5 kb . Incase , both a true positive marker as well as promoter lay within this window , the criteria used to decide if the enhancer was “validated” or “misclassified” is discussed in detail in the Methods section . In H1 cells , we obtained a total of 55382 predicted enhancers at the lowest voting cutoff of 0 . 5 . Over 80% of these predicted enhancers overlap with distal DNase-I hypersensitive sites and the binding sites of p300 , NANOG , OCT4 and SOX2 . Upon randomly generating enhancer predictions in the H1 genome 100 times , we found the average validation rate to be 18 . 43% and the actual validation rate of 80% to be highly significant with a one-sided t-test p-value of 10∧-256 . Additionally , we found that 5% of them overlap with UCSC TSS , indicating a low misclassification rate of 5% ( fig . 2C , E , in red ) . A similar high level of validation rate and low misclassification rate were observed when RFECS was applied to IMR90 cells , where 83581 enhancers were predicted with a validation rate of 85% ( average random validation rate = 16 . 13% , pvalue = 2×10∧-279 ) , and misclassification rate of 4% ( fig . 2D , F ) . Thus , RFECS appears to accurately predict putative enhancer sequences based on chromatin modification state of the genome . We next tried to assess the linear resolution of RFECS predictions . We calculated the distance between the predicted enhancers and locations of enhancer markers such as DNase-I hypersensitive sites , or p300 binding sites in each cell type , and found that the majority of predicted enhancers are within 200 bp of these sites ( fig . S3A , B ) . In H1 , nearly 62% of enhancers lie within 200 bp of an enhancer marker site ( fig . S3A ) , while in IMR90 this value is around 70% ( fig . S3B ) . Thus , the majority of enhancer predictions also show a high distance resolution in terms of proximity to the validation marker . We also confirmed that our enhancer predictions showed an activation of gene expression in the proximal TSS . In order to do this , we compared RNA-seq datasets ( Wei Xie et al . , manuscript under revision ) in H1 and IMR90 using edgeR [37] to identify H1-specific and IMR90-specific TSS . Then we identified enhancer predictions specific to either H1 or IMR90 using a filter distance of 2 . 5 kb . When we look at the average distribution of H1-specific enhancers they are clearly enriched in the vicinity of H1-specific TSS as compared to either non-specific TSS or IMR90-specific TSS ( fig . S3C ) and this enrichment is found to significant at distances up to at least 500 kb using a Wilcoxon test ( p-value<10∧-6 ) . Similarly , in the case of IMR90-specific enhancers , we observe them to be more enriched in the proximity of IMR90-specific TSS as compared to H1-specific TSS ( fig . S3D , p-value<10∧-23 ) . As further evidence that RFECS accurately predicts enhancers , chromatin modifications at the predicted enhancers showed presence of all chromatin states observed in the training sets comprised of a subset of distal p300 binding sites ( fig . 1 ) . In H1 , clusters 1 , 2 and 8 of enhancer predictions ( fig . S4 ) are similar to clusters 1–3 of the p300 binding sites ( fig . 1A ) , clusters 3–4 appear to correspond to cluster 5 of p300 BS , while clusters 5–6 look like cluster 4 of p300 BS . In IMR90 , similar trends could be observed when comparing chromatin states at enhancer predictions ( fig . S5 ) to those of p300 binding sites ( fig . 1B ) . Further , it can be observed that clusters 3–6 of the enhancer predictions in H1 ( fig . S4 ) that have weaker acetylation and/or enrichment of H3K27me3 also tend to have lower voting percentage of trees . In summary , we showed that RFECS accurately predicted enhancers in the two cell lines H1 and IMR90 using a set of 24 chromatin modifications . These enhancers showed high validation rates , low misclassification rates and sharp linear resolution . To make enhancer predictions , our approach requires a construction of a random forest trained on promoter-distal p300 binding sites . It is time-consuming and expensive to create a new training set for enhancer prediction in each new cell type , so it is desirable to use a random forest developed in one cell type to predict enhancers in another . To evaluate the feasibility of such approach , we first trained a random-forest using chromatin modification profiles obtained in H1 , and then applied it to the IMR90 cells . Compared to RFECS predictions using IMR90 chromatin profiles as training set , RFECS predictions using H1 training dataset reduces the validation rate by ∼5–8% and increases the misclassification rate by ∼2% ( fig . 2C , E black vs red ) . Similarly , we also developed a random forest using the IMR90 data as the training set and then applied it to H1 . This led to an average reduction of 2–3% in validation rate ( fig . 2D , black vs red ) . Therefore , RFECS trained using one cell type may be applied to a different cell type , albeit with slightly lower accuracy . We sought to examine if this moderate decrease in performance was largely due to cell-type specific differences or was within the limits of technical or biological variability between replicates . To this end , we trained a random forest on one replicate of a cell-type , and made predictions on the other replicate of the same cell type . RFECS trained on IMR90 and then applied to the replicate 1 of the H1 profiles ( blue dot vs asterisk ) actually showed a higher validation rate and lower misclassification rate than RFECS trained using replicate 2 of H1 ( fig . 2C , E ) , while similar performance was observed with enhancer predictions on replicate 2 of H1 independent of whether the random-forest was trained on H1 replicate 1 or IMR90 ( green dot vs asterisk ) . Similar trends were observed when comparing predictions made on individual replicates of IMR90 using either H1-training or training on the other replicate ( fig . 2D , F ) . In conclusion , predicting enhancers using the random forest built from a different cell type exhibits a modest decrease in performance compared to a same-cell training set . However , this decrease in performance is comparable to the decrease that can arise due to variability between two replicates of the same cell-type . With the increasing number of histone modifications being discovered and mapped , determination of the relative importance of each mark in defining genomic elements is important . An out-of-bag measure of variable importance is a natural by-product of random forest classification scheme [33] wherein the relative importance of each feature is assessed as the increase in classification error upon permutation of feature values across classes . In both H1 and IMR90 , the variable importance was assessed for random forests trained on 5 cross-sections of data for each of the 2 sets of replicates individually as well as the set of averaged replicates . Upon ranking histone modifications by variable importance , it is apparent that H3K4me1 and H3K4me3 are the top 2 most robust modifications across replicates and cross-sectional samples in both cell types , followed by H3K4me2 ( fig . 3A , B ) . This indicates that these 3 modifications maybe the most informative in the prediction of enhancers in any unknown cell type as well . Beyond the top 3 modifications , there is variability among the cell types . In IMR90 , the other modifications appear to contribute almost equally , while in H1 there is a much clearer difference in variable importance . These differences are supported by correlation analyses in H1 and IMR90 ( fig . 3C , D ) . In H1 , several modifications are highly correlated , which could explain the larger differences in variable importance , as only a few variables maybe needed to form a non-redundant set . In IMR90 , the correlations are lower and hence each of the modifications may contribute non-redundant information and thus contribute equally to the variable importance . Modifications that cluster together in both H1 and IMR90 ( shown in the same non-black colors , fig . 3C , D ) suggest cell-type independent redundancy . Having established the relative importance of each histone modification in predicting enhancers , we next examined the accuracy of predictions using different sets of modifications . Validation rates obtained by using the minimal set of H3K4me1-3 is within 2% of that for all 24 modifications in H1 ( fig . 4A ) . Furthermore , this minimal set performs considerably better than the more conventionally selected set of H3K4me1 and H3K4me3 [3] , [5] and at times , H3K27ac [38] , [39] ( fig . 4A , B , in black and blue ) . The set of H3K4me1-2-3 is more comparable to H3K4me1-H3K4me3-H3K27ac in IMR90 but does have a slightly lower misclassification rate ( fig . 4D ) . In both cases the use of the minimal set of 3 modifications shows a much closer resemblance in performance to all 24 modifications than to the set of 2 marks H3K4me1 and H3K4me3 ( fig . 4A–D ) . It can also be observed that in conjunction with H3K4me1 and H3K4me3 , using H3K4me2 picks up a larger proportion of enhancers with weaker acetylation enrichment as compared to H3K27ac ( fig . S4 , S5 ) , supporting our prediction of the minimal set . We also made enhancer predictions using all possible combinations of 3 modifications in chromosome 1 for replicate 1 and replicate 2 of H1 . The average validation rate for a fixed range of enhancers was compared across replicates and it can be seen the set corresponding to H3K4me1 , H3K4me2 and H3K4me3 ( marked in * ) , is the highest performing combination common to both replicates ( fig . 4E ) . We also found the performance of the combination of H3K27ac with H3K4me1 and H3K4me3 appears to be comparable in this case ( 3 , fig . 4E ) , validating the use of H3K27ac as a feature for enhancer prediction when H3K4me2 is not available . Some of the worst performing combinations include H3K9me3 and H4K20me1 ( 4 and 5 , fig . 4E ) , which also show up as variables with least importance in fig . 3A . In many currently existing datasets , H3K27ac is the more commonly sequenced histone modification as compared to H3K4me2 due to it's perception as a marker of active enhancers . While using H3K4me2 may improve enhancer prediction in some cell-types , use of H3K27ac in addition to H3K4me1 and H3K4me3 marks does show considerable improvement over using just the top 2 marks H3K4me1 and H3K4me3 ( fig . 4A–D ) . Hence , for many of the currently existing datasets , we could use H3K4me1 , H3K4me3 and H3K27ac as the features in our random-forest with satisfactory performance . Overall , these comparisons indicate the suitability of selecting H3K4me1 , H3K4me2 and H3K4me3 as three minimal chromatin marks for purposes of enhancer prediction . Additional chromatin modifications required for improving upon enhancer predictions may depend on cell-type specific characteristics , as indicated by the differences in variable importance between H1 and IMR90 ( fig . 3A , B ) . We next asked if our enhancer prediction algorithm performed better than several other current techniques for enhancer prediction – CSIANN , ChromaGenSVM and Chromia [23] , [24] , [26] , [39] . In previous studies , CSIANN and ChromaGenSVM were applied on the histone modification dataset in CD4 T-cells [24] , [26] , [39] . In order to make a comparison of performance of our method with previous approaches , we applied RFECS to the CD4+ T cell dataset as well and determined parameters using cross-validation ( fig . S6 ) . Using H3K4me1 , H3K4me3 , and H3K27ac , CSIANN made 21832 predictions [39] and ChromaGenSVM method made 23574 predictions [26] . We made enhancer predictions using H3K4me1 , H3K4me3 and H3K27ac with RFECS as well as Chromia [23] . Cutoffs were selected that yielded a similar number of enhancer predictions for both Chromia ( 21895 ) and RFECS ( 22947 ) ( fig . 5A ) , so as to make a fair comparison across methods . To compare these different sets of enhancer predictions , we computed validation rates by comparing them to TSS-distal DNase-I hypersensitive sites , p300 binding sites , and CBP binding sites and misclassification rates by comparing to known UCSC TSS using a window of −2 . 5 kb to +2 . 5 kb as described in the methods . ( fig . 5A ) . The validation rate of RFECS predictions is around 70% , which is considerably higher than the other three methods ( 57% ChromaGenSVM , 51% CSIANN , 60% Chromia ) . Further , the misclassification rates of RFECS is less than 7% , much lower than the 27% , 35% and 15% rates of ChromaGenSVM , CSIANN and Chromia , respectively . These results suggested that overall procedure for RFECS , including selection of training set as well as training and prediction using the vector-random-forest , performs better than currently available techniques for enhancer prediction . In the above comparison , we selected our enhancer-representative training set as p300 peaks called using MACS [40] that were distal to known UCSC TSS and overlapped DNase-I locations while CSIANN and ChromaGenSVM used a training-set of p300 peaks called using SICER previously [41] . We also wanted to compare the performance of the different algorithms on our own datasets using the same training-set to evaluate the performance of the random-forest based part of the algorithm . To achieve this , we ran the various enhancer prediction methods on H3K4me1 , H3K4me2 and H3K4me3 datasets of H1 , with help from the author of ChromaGenSVM [26] ( fig . 5B ) . We tried to make the pre-processing stages of the various algorithms as consistent as possible by merging several replicates of each histone modification files and input files into single bed files and randomly selecting a smaller subset of p300 peaks for training , since these were the requirements of the other algorithms such as CSIANN and ChromaGenSVM . Incase of CSIANN , the selection of background was hard-coded in the software but in all other cases we used our own background training set as well . In fig . 5B , it can be observed that RFECS shows a maximum validation rate of around 82 . 8% as compared to 66 . 8% , 57 . 7% and 63 . 3% for ChromaGenSVM , CSIANN and chromia respectively . Further , RFECS showed the lowest misclassification rate of 4 . 9% as compared to 8 . 3% , 36 , 7% and 10 . 1% rates for the above-mentioned cases . Hence , the improvement in performance due to RFECS cannot be solely attributed to method of selecting the training-set . In summary , RFECS shows considerably improved performance over existing enhancer-prediction algorithms in two very different datasets and hence can be considered an advance in the field . Comparing enhancer predictions across diverse cell-types can contribute to understanding differences in regulatory mechanisms between cell-types . The ENCODE dataset is an example of a collection of high-throughput datasets such as histone modifications and transcription factor binding data that are available for multiple cell-types [42] . Having a set of high-confidence enhancer predictions in these cell-types would be a valuable resource . We trained our random forest on the p300 ENCODE data in H1 and made enhancer predictions in 12 ENCODE cell-types using the three marks H3K4me1 , H3K4me3 and H3K27ac since these were available for all the cell-types . Validation rates were assessed based on overlap with existing DNAse-I hypersensitivity data while misclassification rates were calculated based on overlap with UCSC TSS . It can be seen that the majority of cell-types show high validation rates between 80 and 95% , while the misclassification rates lie within acceptable levels of 2–7% ( fig . 6A , B ) . In order to compare enhancers across cell-types , it is preferable to have enhancer predictions with the same level of confidence . To determine the appropriate cutoff for multiple number of cell-types , we calculate a False Discovery rate by randomly permuting 100 bp bins across the genome and computing the ratio of enhancers predicted in permuted data/enhancers predicted in real data for various cutoffs of voting percentages . In fig . 6C , it can be seen that different cell-types show a different relationship with FDR . For example , at an FDR of 5% , the voting percentage for GM12878 ( solid dark blue ) is 0 . 74 , for Nhek ( dashed cyan ) 0 . 64 and for Hsmm ( solid yellow ) it is 0 . 56 . Using an FDR of 5% , we obtained a consistent set of high-confidence enhancer predictions in the 12 ENCODE cell-types . In fig . 6D , the numbers of enhancer predictions in each cell type is shown above the bar . The validation rates ( in red ) are above 90% for all cell-types except H1 , Hepg2 and GM12878 . In H1 and Hepg2 , the numbers of DNase-I hypersensitivity sites are relatively less , i . e . ∼150 to 177K as compared to ∼230 to 380K in the other cell-lines . This may explain the somewhat lower validation rate in these two cell-types . GM12878 appears to be an outlier and we suspect that enhancer predictions may potentially be improved in this cell line by using a different training set . In summary , we obtained a high-confidence set of enhancer predictions in multiple ENCODE cell-lines with the same level of confidence . This will enable more rigorous comparisons of regulatory characteristics of these cell-types in the future .
We describe here a novel machine-learning algorithm to accurately predict enhancers in a genome-wide manner based on chromatin modifications . We trained this algorithm using novel p300 training sets in H1 and IMR90 and 24 chromatin modifications in each cell-type . We showed that models trained on one cell-type could be effectively applied on another cell-type . Random forests enable detection of the most informative features required for a classification task . In the case of enhancer prediction , we identified a set of 3 histone modifications that appeared to be the most informative and robust across cell-types and replicates . Such an approach can once again be applied when the number of genome-wide modification maps is expanded in various different cell types and the most informative set of modifications can be further refined . We show that RFECS outperforms other machine-learning based prediction tools in CD4+ T cells , and can be applied in the future to multiple cell types . We successfully applied our enhancer prediction tool to 12 cell-lines in the publicly available ENCODE database and obtained a set of enhancers with a consistently high level of confidence across the cell-types . In the future , we could potentially adapt the RFECS method to detect other regulatory genomic elements that can be observed to have a distinct chromatin signature and find the minimal set of chromatin marks for this purpose . The ability to detect diverse patterns of features within the training set indicates that the RFECS approach could be used to train on a composite training set comprised of different transcription factors . Combining information from different enhancer-binding proteins may improve prediction of regulatory elements . Random forests are non-parametric and have been shown to integrate a large number of diverse features . This could suggest the addition of other discrete and continuous data types such as sequence or motif based information or DNA methylation to the prediction of genomic elements .
The H1 and IMR90 datasets used in this study were generated as part of the NIH Roadmap Epigenome Project and have been released to the public prior to publication ( http://www . genboree . org/epigenomeatlas/multiGridViewerPublic . rhtml ) . Briefly , 24 chromatin modifications in human embryonic stem cell ( H1 ) and primary lung fibroblast cells ( IMR90 ) were generated by the Ren lab and deposited under the NCBI Geo accession number GSE16256 . Additionally , two replicates of H3K9me3 datasets deposited under Geo accession numbers GSM818057 and GSM42829 were used . Genome-wide binding data for p300 in H1 and IMR90 , and transcription factors NANOG , SOX2 and OCT4 in H1 were generated in the Ren lab using ChIP-seq and deposited under accession numbers GSE37858 , GSE18292 and GSE17917 respectively . Any data mapped to hg18 was converted to hg19 using liftover tools [43] . The DNase-I hypersensitivity datasets for H1 and IMR90 were produced by the Stammatoyanopoulos group at UW [44] . IMR90 DNase-I raw data may be accessed using GSM468792 and narrow peak calls are attached as supplemental information . Narrow DNase-I peaks in H1 were downloaded from UCSC ENCODE page ( http://hgdownload . cse . ucsc . edu/goldenPath/hg19/encodeDCC/wgEncodeUwDnase/ ) For CD4 , previously generated datasets for p300 [41] , CBP [41] and DNase-I [21] data as well as histone modifications [45] , [46] were used . Histone modification data and DNase-I hypersensitivity data for the 12 ENCODE cell-lines was downloaded from http://genome . ucsc . edu/ENCODE/downloads . html . The ChIP-seq reads for the histone modification as well as corresponding input were binned into 100 bp intervals . The binned modification file was normalized against the binned input file using an RPKM ( Reads per kilobase per million ) measure [47] . In the case of 2 or more replicates , the RPKM- level for each bin is averaged to get a single histone modification file , in order to minimize batch-related differences . MACS [40] software was used to call peaks for p300 , CBP and any other TF such as NANOG , SOX2 and OCT4 . ChIP-seq input files were used as background and parameters of mfold = 20 and default p-value cutoffs were used . Peak calls are available as supplemental files . In case of the p300 and CBP binding sites used to validate enhancer predictions in CD4 , we included the regions of enrichment that were previously published as well [41] We constructed the forest using the concept of binary classification trees , with each feature being a 20-dimensional vector of 100 bp bins from −1 to +1 kb along the genomic element . At each node in the tree , a linear classifier was constructed using the Fischer Discriminant approach using the histone modification vector , allowing for utilization of shape as well as abundance information ( fig . S7A ) . The utilization of the linear discriminant at each node was inspired by the recent development of methods such as the discriminant random-forests [34] and oblique random forests [35] . The Vector-Random forest algorithm was implemented in MATLAB ( MATLAB 7 . 14 . 0 . 739 , The Mathworks Inc . , Natick , MA , 2012a ) as the function “multiclasstree” and utilizes functions from the “classregtree” and “classify” functions of MATLAB , implementing decision trees and linear discriminants respectively . The code used for RFECS can be downloaded from: http://enhancer . ucsd . edu/renlab/RFECS_enhancer_prediction/ Enhancer prediction involved two stages , which are classification of p300 vs non-p300 and peak-calling . A major advantage of the random forest is the inherent ability to select more important variables versus less important ones . In order to compute the order of variable importance , in this case , the importance of individual histone modifications for making enhancer predictions , we use an out-of-bag measure of variable importance [33] implemented in Matlab as the function oobVarImp . Based on the ordering of the variable importance across 5 different cross-sections of the training dataset of multiple replicates and cell types , certain modifications may always be observed to have priority . Due to the non-redundant nature of the ordering of variables as well as their robustness across replicates and samples , these modifications maybe selected as the most informative ones that are required to make enhancer predictions . Cross-validated ROC curves were used to estimate parameters for use within the same algorithm . However , comparisons across different algorithms may be biased depending upon the composition of the training set , so we validated enhancer predictions as described below . Enhancer Predictions outputted from the random forest predictor have background enrichment scores of “voting percentage” ranging from 0 . 5 to 1 to enable detection of enhancers at different levels of confidence . At higher cutoffs , confidence of prediction is higher , but fewer enhancers are detected . The availability of large-scale datasets such as DNase-I hypersensitive sites , p300 binding sites , CBP binding sites and transcription factor binding sites enabled an estimate of the number of true positives at every cutoff . Further , the number of enhancers misclassified as TSS at each cutoff was also determined . Within the same cell type , an enhancer prediction method that performs better , should pick up more true positive validation markers and fewer TSS , given the number of predictions are the same . Predicted enhancers are classified as “validated” , “misclassified” or “unknown” based on the criteria below . True Positive Markers ( TPM ) refer to DNase-I hypsersensitivity site , p300 , CBP and Transcription factor binding sites . The Pearson correlation coefficient between any two modifications was computed for RPKM-normalized histone modification reads between −1 to +1 kb for all elements within the selected training set . The correlation patterns of each histone modification was used to cluster the modifications and order them using MATLAB tools . This enabled visualization of which modifications are the most similar in their correlation patterns . In the ordering of variable importance , if certain variables showed up as important in two different cell types , the redundancy based on their correlation plots could be used to explain away this variability . ChromaSig [48] was used to cluster histone modification patterns along p300 binding sites and predicted enhancers using modification width as 4 kb . The resulting clusters were then visualized using Java TreeView [49] .
|
Enhancers are regions in the genome that can activate the expression of a gene irrespective of their location with respect to the gene . Identifying these elements is critical in understanding regulatory differences between different cell-types . Since enhancers lack characteristic sequence features and can be far away from the gene they regulate , their identification is not trivial . Experimentally determining the genome-wide binding sites of transcriptional co-activator p300 is one way of finding enhancers but it can only identify a subset of enhancers . A few years ago , it was observed that the binding sites of p300 are marked by distinctive , post-translational histone modifications . Several groups have exploited this discovery to predict genome-wide enhancers based on their similarity to the histone modification profiles of p300 binding sites . We here report a novel algorithm for this purpose and show that it has much greater accuracy than existing methods . Another unique feature of our algorithm is the ability to automatically deduce the most informative set of histone modifications required for enhancer prediction . We expect that this method will become increasingly useful with the expanding number of known histone modifications and rapid accumulation of epigenomic datasets for various cell types and species .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"computer",
"science",
"biology"
] |
2013
|
RFECS: A Random-Forest Based Algorithm for Enhancer Identification from Chromatin State
|
Microorganisms in nature do not exist in isolation but rather interact with other species in their environment . Some microbes interact via syntrophic associations , in which the metabolic by-products of one species serve as nutrients for another . These associations sustain a variety of natural communities , including those involved in methanogenesis . In anaerobic syntrophic communities , energy is transferred from one species to another , either through direct contact and exchange of electrons , or through small molecule diffusion . Thermodynamics plays an important role in governing these interactions , as the oxidation reactions carried out by the first community member are only possible because degradation products are consumed by the second community member . This work presents the development and analysis of genome-scale network reconstructions of the bacterium Syntrophobacter fumaroxidans and the methanogenic archaeon Methanospirillum hungatei . The models were used to verify proposed mechanisms of ATP production within each species . We then identified additional constraints and the cellular objective function required to match experimental observations . The thermodynamic S . fumaroxidans model could not explain why S . fumaroxidans does not produce H2 in monoculture , indicating that current methods might not adequately estimate the thermodynamics , or that other cellular processes ( e . g . , regulation ) play a role . We also developed a thermodynamic coculture model of the association between the organisms . The coculture model correctly predicted the exchange of both H2 and formate between the two species and suggested conditions under which H2 and formate produced by S . fumaroxidans would be fully consumed by M . hungatei .
Microorganisms in nature engage in a variety of interactions with other species in their environment . Syntrophy is one such type of inter-species interaction in which one species lives off the metabolic by-products of another [1–3] . Synthetic methanogenic communities [4] are typically tightly constrained by thermodynamics , as the oxidation reactions carried out by the first community member are thermodynamically unfavorable unless the degradation products are maintained at low levels by the second community member [5] . In anaerobic syntrophic communities , electrons are transferred from one partner to the other through direct contact or small molecule diffusion [6] . Traditional biochemistry has elucidated intracellular electron transport mechanisms [3 , 7–9] , but it is difficult to evaluate these pathways in their metabolic and environmental context . Genome-scale metabolic models ( GEMs ) [10–12] and constraint-based methods are powerful computational tools for understanding individual pathways in a broader metabolic context including both isolated microbial species [13–15] and simple microbial communities [16–26] . One of the earliest microbial community models used flux balance analysis ( FBA , [27] ) to investigate formate and H2 exchange between the sulfate-reducing bacterium Desulfovibrio vulgaris and the methanogenic archaeon Methanococcus maripaludis [16] . In this study , each organism was modeled as a compartment within a larger community-scale model . Compartmentalized approaches have been used to study the origins of cooperation and competition [17–19] , as well as specific communities [20–24] . These approaches [16–24] have often used a single ( joint ) objective function to capture community behavior . OptCom [25 , 26] instead uses a multi-level optimization framework , to capture the trade-offs between individual and community fitness , with separate objective functions for the individual species and the community . In addition , community FBA ( cFBA ) [28] extends compartmentalized approaches [16–24] to specifically account for individual species’ biomass abundance . Genome-scale models can also be used to study the relationship between thermodynamics and metabolism , by ensuring that network predictions are consistent with thermodynamic principles [29–34] . In this study , we used thermodynamics-based metabolic flux analysis ( TMFA ) to develop a thermodynamic , coculture model of the syntrophic association between the anaerobic bacterium Syntrophobacter fumaroxidans and the methanogenic archaeon Methanospirillum hungatei . In association with M . hungatei , S . fumaroxidans converts propionate to acetate , CO2 , and H2 [35–37] . CO2 and H2 can be interconverted to formate [38–40] , with H2 and formate serving as the electron carriers between the two species . H2 and formate production are only observed during syntrophic growth . Using a thermodynamic , constraint-based model , we set out to test the proposed hypothesis that this behavior is governed by thermodynamics [3 , 5 , 6 , 8 , 9] . We developed genome-scale metabolic reconstructions of both microorganisms , and verified proposed mechanisms of ATP production within each individual species . Additional constraints and a cellular objective function were identified to predict the proper flux through experimentally characterized carbon and electron transport pathways during monoculture and syntrophic ( i . e . , coculture ) growth . Our analysis revealed that thermodynamic constraints alone are insufficient to explain why S . fumaroxidans does not produce H2 in monoculture . We also extended TMFA to model the syntrophic association between the two microorganisms . The association is modeled as a continuous coculture system with constraint-based models for each microbe and a mass balance around the reactor . Similar to cFBA [28] , the coculture model accounted for the biomass concentrations of each species . We predicted the behavior of this syntrophic association under a variety of dilution rates , and identified regimes of behavior consistent with experimental observations .
The iMhu428 reconstruction of M . hungatei was built from the iMB745 reconstruction of M . acetivorans [41] . A preliminary draft reconstruction was built from iMB745 using the RAVEN Toolbox [42] and the KEGG SSDB [43]; however , M . hungatei orthologs were found for only 428 of the 745 genes in iMB745 . To avoid extensive gapfilling , reactions from the iMB745 were copied into the M . hungatei reconstruction , with modifications to reflect key metabolic features of M . hungatei ( see S1 Text ) . As a consequence , the iMhu428 reconstruction is a draft reconstruction requiring further evaluation . A thermodynamic model for iMhu428 was built and TMFA was used to predict ATP generating mechanisms in minimal media monoculture conditions ( see S1 Table in S1 Dataset for constraints used ) . Experimental evidence suggests that M . hungatei is able to generate 0 . 5 mole ATP per mole of CO2 converted to CH4 [7] , via the metabolic route shown in Fig 1 . In order for this route to be thermodynamically feasible , the ΔrG'0 of one reaction ( FMFTSPFT , formylmethanofuran-tetrahydromethanopterin formyltransferase ) had to be allowed to vary within a 99% confidence interval of its estimated standard transformed Gibbs free energy of reaction ( ΔrGest'0 ) ( rather than the 95% interval used for all other reactions ) in order to carry flux in the proper direction . The reactions for carbon source utilization in M . hungatei are well-characterized [7 , 44–47] , but uncertainty remains about the stoichiometry of small ion transport [7] . Na+ transport stoichiometries associated with tetrahydromethanopterin S-methyltransferase ( MTSPCMMT_CM5HBCMT , E . C . 2 . 1 . 1 . 86 ) and a Na+/H+ antiporter ( NAT3_1 ) were selected to give an ATP yield matching the experimental estimates: two Na+ ions exported by MTSPCMMT_CM5HBCMT , and a one Na+ per H+ transported by NAT3_1 . However , different stoichiometries for these reactions are also thermodynamically possible ( see Discussion ) . Experimental measurements of growth rates , yields , and maintenance costs were also used to identify substrate uptake rates ( SUR ) for CO2 and formate , and the growth- ( GAM ) and non-growth-associated ( NGAM ) ATP maintenance requirements for M . hungatei , as described in S1 Text . NGAM represents the amount of energy spent to maintain the cell ( i . e . , maintenance energy ) , while GAM represents energy spent on growth-related functions ( e . g . , protein synthesis ) . For the iMhu428 model , the NGAM was estimated to be 0 . 6 mmol ATP/gDW/day , GAM was estimated to be 47 mmol ATP/gDW , SURCO2 was estimated to be 75 . 7 mmol/gDW/day , and SURformate was estimated to be 955 mmol/gDW/day . The iSfu648 reconstruction of S . fumaroxidans was built from the KEGG database using the RAVEN Toolbox [42] . The resulting draft reconstruction was manually refined ( see S1 Text ) , with particular attention paid to ATP production mechanisms . A number of studies have identified gene clusters encoding a variety of hydrogenases , dehydrogenases , and other electron transport enzymes [8 , 9 , 48–51] , whose expression levels vary across growth conditions [51] . All told , 17 enzymes which catalyze 12 different electron transport reactions have been identified ( S3 Table in S1 Dataset ) . In many cases , the proposed reactions catalyzed by these enzymes differ between studies; a brief description of each reaction and justification for each annotation is given in S1 Text . The draft reconstruction was updated to be consistent with the reported carbon utilization and electron transport reactions , and the resulting stoichiometric model was converted to a thermodynamic model . Experimental studies have elucidated five growth modes for S . fumaroxidans: four in monoculture and one in coculture with M . hungatei ( S1 Text ) [36 , 48 , 52] . This work examines the three most commonly studied growth modes ( Table 1 ) : monoculture growth on fumarate , monoculture growth on fumarate plus propionate , and coculture growth on propionate . A variety of experimental findings were synthesized to develop theoretical flux distributions for these three growth modes ( Fig 2 and S1 Text ) . These experimental findings suggested additional regulatory and flux-coupling constraints for the iSfu648 model , such as coupling between fumarate reductase and the cytosolic hydrogenase due to co-localization in the membrane ( see S2 Table in S1 Dataset and S1 Text for the full set of constraints and their justification ) . During monoculture growth on fumarate alone ( Fig 2A ) , one mole of fumarate gets fully oxidized to CO2 , while six moles of fumarate get reduced to succinate [48 , 52]: 7fumarate→6succinate+4CO2 ( 1 ) The oxidation of one fumarate to CO2 generates one ATP and five reducing equivalents ( three NADH and two pairs of reduced ferredoxin ) [48 , 52] , while the reduction of additional fumarate to succinate by fumarate reductase ( FRD ) consumes reducing equivalents ( menaquinol ) [48] . Electrons are transferred from NADH and reduced ferredoxin to menaquinone through the combined action of the Rnf complex ( RNF ) , the ferredoxin-oxidizing hydrogenase ( frH2ase ) , the cytosolic hydrogenase ( cytH2ase ) and formate hydrogen lyase ( FHL ) . The reduction of fumarate to succinate also generates the proton motive force ( PMF ) responsible for driving the RNF reaction and producing ATP . During monoculture growth on fumarate plus propionate ( Fig 2B ) , one mole of propionate gets oxidized to succinate , while one mole of fumarate gets oxidized to acetate and CO2 . Two additional moles of fumarate get reduced to succinate [36 , 48]: propionate+3fumarate→acetate+CO2+3succinate ( 2 ) The oxidation of fumarate to acetate and CO2 produces one NADH and one pair of reduced ferredoxin , while the reduction of fumarate to succinate by FRD consumes menaquinol . Electrons are transferred from NADH and reduced ferredoxin to menaquinone through the combined action of the confurcating hydrogenase ( cH2ase ) and cytH2ase . Oxidation of propionate to succinate produces one ATP , while FRD generates the PMF necessary for additional ATP production . During coculture growth on propionate ( Fig 2C ) , propionate gets oxidized to acetate and CO2 via the methylmalonyl-CoA pathway [48 , 52]: ATP is generated during the oxidation of propionate to succinate , and this ATP establishes the PMF necessary to drive the endergonic oxidation of succinate to fumarate ( SDH ) , producing menaquinone . cytH2ase then transfers electrons from menaquinol to two protons , generating H2 . The oxidation of fumarate to acetate and CO2 produces one NADH and one pair of reduced ferredoxin , and cH2ase couples NADH and ferredoxin re-oxidation with H2 production . Unlike in the monoculture growth modes , the H2 is not consumed intracellularly and must diffuse outside the cell . It has been proposed that the net production of H2 by S . fumaroxidans is only thermodynamically favorable at the low H2 concentrations maintained by methanogens , thereby explaining why S . fumaroxidans only produces H2 during coculture growth . S . fumaroxidans exhibits considerable flexibility in its ATP production mechanisms during coculture growth ( [8 , 9 , 48 , 51] ) , and can produce formate instead of CO2 ( Fig 3 ) yielding an overall transformation of: propionate→acetate+formate+2H2 ( 4 ) In one mechanism ( Fig 3A ) , activity of the cytosolic formate hydrogenase ( cytFDH ) substitutes for the activity of cytH2ase . In a second mechanism ( Fig 3B ) , the confurcating formate dehydrogenase ( cFDH ) substitutes for cH2ase . Here , cFDH couples NADH and ferredoxin re-oxidation with the conversion of CO2 ( from propionate oxidation ) to formate . Experimental evidence and conceptual models of S . fumaroxidans energy metabolism suggest that the carbon and electron transfer pathways shown in Fig 2 provide the sole source for ATP production in S . fumaroxidans , either by substrate-level phosphorylation or through establishment of a proton gradient used by ATP synthase [5] . To test the computational model’s predictions , TMFA was used to maximize ATP production under each of the three growth modes . When flux was restricted to a reduced network containing all the reactions shown in Fig 2 ( listed in S4 Table in S1 Dataset ) , the iSfu648 model correctly predicted the flux distributions shown in Fig 2 . However , when flux was allowed throughout the entire network , additional flux distributions with higher ATP yields were identified . Additional reaction direction constraints were developed to ensure model-predicted flux distributions matched experimental observations ( see S2 Table in S1 Dataset and S1 Text for details ) . However , the resulting flux distributions are not fully consistent with the hypothesis that S . fumaroxidans has adapted to maximize its energy yield . Experimental studies of S . fumaroxidans have shown that H2 is not produced during growth in monoculture [36 , 52 , 53] , and it is widely thought that H2 production is only thermodynamically favorable at low partial pressures [3 , 6 , 8 , 9] . In particular , methanogens in syntrophic communities enable sustained H2 production by consuming H2 and keeping its partial pressure low [3 , 5 , 6 , 8 , 9] . Indeed , when H2 production was observed in monoculture , H2 production ceased at a partial pressure of approximately 10 Pa [53] . However , when maximizing H2 production under monoculture conditions , simulations reveal that H2 production remains thermodynamically feasible . For example , during monoculture growth on fumarate , the iSfu648 model predicts that H2 can be produced via the following mechanism: fumarate→4CO2+6H2 ( 5 ) In this scenario , H2 molecules produced by the ferredoxin-oxidizing hydrogenase are exported outside the cell , instead of serving as substrates for the cytosolic hydrogenase . As a result , no PMF is generated by fumarate reductase , and the net ATP yield is zero . Thus , while H2 production remains thermodynamically possible , H2 production is only associated with sub-optimal mechanisms of ATP generation . This suggests that thermodynamic considerations alone may not explain the absence of H2 production during monoculture growth , but that the observed flux distribution may instead be driven by demands for energy generation . While H2 production was not predicted for monoculture conditions when ATP production was maximized , H2 production was initially predicted when growth was instead maximized . To eliminate monoculture H2 production in the TMFA model , we first sought to constrain ratios of metabolite concentrations with an approach similar to that used to correct TMFA growth predictions [34] ( see S1 Text for details ) . While metabolite ratio constraints could be identified to prevent some H2 production mechanisms , H2 production during monoculture growth could not be completely eliminated . If thermodynamics prevents H2 production in monoculture , then the current thermodynamic model may contain too much uncertainty in its Gibbs free energy estimates ( see Discussion ) . Regulatory effects could also potentially prevent H2 production in monoculture conditions . To correct the model , all subsequent monoculture simulations were performed by preventing H2 production . Model parameters were estimated after reaction direction constraints were added to the iSfu648 model ( to be consistent with reported ATP generation and H2 production mechanisms ) . Experimental measurements of growth rates , yields , and maintenance costs were used to identify the SURs , GAM , and NGAM parameters for S . fumaroxidans . These parameters were estimated using data from monoculture growth on fumarate alone and coculture growth on propionate alone ( see S1 Text ) . For the iSfu648 model , the following parameters resulted in the best fit of the model to the experimental data: NGAM = 3 . 36 mmol ATP/gDW/day , GAM = 22 . 8 mmol ATP/gDW , SURpropionate = 37 . 7 mmol/gDW/day , and SURfumarate = 27 . 6 mmol/gDW/day . Using these parameter values , the in silico growth rates under each growth condition were predicted ( Table 1 ) . Not surprisingly , the predicted growth rates for fumarate alone and propionate alone conditions agree with experimental observations ( since these were used to estimate the parameter values ) . However , the model significantly under-predicts the measured growth rate during monoculture growth on fumarate plus propionate ( 0 . 55 days-1 predicted , 0 . 73 days-1 observed ) . This discrepancy could be caused by differences in uptake rates or maintenance costs in the fumarate plus propionate condition compared to the conditions with propionate alone or fumarate alone . When maximizing biomass production on the entire network , the iSfu648 model predicted a wide range of product secretion rates . When the enzyme cost ( i . e . , total flux ) was minimized at the maximum growth ( pTMFA [54] , see Methods ) the model-predicted product yields closely matched reported values for two of the three growth modes ( Table 1 ) —monoculture growth on fumarate alone and coculture growth on propionate alone . These results indicate that the majority of carbon is diverted to fermentation products , consistent with the expectation that high fluxes through the low-energy fermentation pathways are needed to meet cellular energy demands . However , for monoculture growth on fumarate plus propionate , the model failed to predict that fumarate and propionate should be consumed at the observed ratio of approximately three fumarate per propionate [36] . Instead , the model predicted both substrates would be consumed at their maximum SURs , resulting in a ratio of 0 . 73 fumarate per propionate . The experimentally observed 3:1 ratio is thought to arise due to coupling within the metabolic network ( Fig 2B ) , since oxidation of one fumarate produces one CO2 ( used to oxidize one propionate ) and two pairs of electrons ( used to reduce two fumarate ) . While this coupling arises naturally on the reduced network , the full metabolic network enables alternative coupling mechanisms ( not shown ) that permit other fumarate to propionate ratios . Since propionate oxidation generates carbon precursors and ATP for biomass , maximizing biomass production results in the model under-predicting the fumarate to propionate uptake ratio . Increasing the ratio of fumarate to propionate uptake ( to 3:1 ) decreases the predicted propionate SUR and growth rate ( S1 Fig in S1 Text ) , implying the experimental ratio is sub-optimal with respect to growth maximization . Instead of constraining SURs ( since values were not reported in the literature ) , we incorporated a fumarate to propionate SUR ratio constraint for this condition . While the predicted product yields closely matched experimental observations ( after imposing the SUR ratio constraint ) , the pTMFA-predicted intracellular flux distribution during monoculture growth on fumarate and propionate substantially deviated from that shown in Fig 2B . Further constraints on reaction directions were required so that propionate and fumarate were metabolized in the model via the pathways shown in Fig 2B ( results not shown ) . Taken together , the need for constraints on reaction directions and fumarate to propionate uptake ratio suggests that neither maximization of biomass nor minimization of enzyme cost are sufficient to explain the fluxes of S . fumaroxidans growing in this monoculture condition . During growth in coculture , S . fumaroxidans converts propionate to acetate , H2 , and CO2 or formate [36 , 52] , while M . hungatei consumes acetate , CO2 , H2 , and formate and produces CH4 [55] . M . hungatei can also optionally interconvert excess CO2 and H2 to formate via a formate dehydrogenase [56] . Cocultures of M . hungatei and S . fumaroxidans have been grown in both batch and continuous ( chemostat ) systems . During batch growth , H2 pressure rose during the lag phase and became constant during exponential growth [53] . Continuous cultures also exhibited constant H2 partial pressure [53] . However , to the best of our knowledge , measurements for the relative ratios of M . hungatei to S . fumaroxidans at constant H2 pressure ( where H2 consumption and production rates are balanced ) have not been reported . Instead , an overall reaction for the coculture of propionate→acetate+0 . 25CO2+0 . 75CH4 ( 6 ) is frequently discussed [5 , 53] , which can occur at a ratio of three M . hungatei to four S . fumaroxidans . Since initial conditions for batch experiments were not reported , a continuous culture model was constructed and first evaluated using a 3:4 relative biomass ratio ( M . hungatei: S . fumaroxidans ) . The continuous coculture model included constraint-based models for each microbe and mass balances around the reactor ( S2 Fig in S1 Text ) , and accounted for the biomass concentrations of each species . Both species were constrained to grow at the dilution rate , and the model minimized the species-weighted total flux through the two metabolic networks ( pTMFA ) . Propionate was the only substrate in the reactor feed ( i . e . , it had a net flux into the reactor ) , ensuring that all carbon and electrons used by M . hungatei were produced by S . fumaroxidans . Predicted yields around individual species were first evaluated ( Fig 4A ) . At low dilution rates , S . fumaroxidans was predicted to convert propionate to acetate , H2 , and CO2/formate and M . hungatei was predicted to convert CO2/formate to CH4 ( Fig 4B and 4C , which show alternate solutions with CO2 or formate being exchanged ) . As the reactor dilution rate increased , the predicted species yields of H2 and CO2/formate ( S . fumaroxidans ) and CH4 ( M . hungatei ) decreased slightly ( Fig 4B and 4C ) . In addition , species’ uptake rates of propionate and CO2/formate increased with dilution rate , as biochemical transformation of these substrates provides the energy needed for cellular growth and maintenance . Species’ uptake rates , secretion rates , and relative biomass ratio affects overall bioreactor yields , and these bioreactor yields were subsequently investigated ( Fig 4D ) . At a 3:4 relative biomass ratio ( M . hungatei: S . fumaroxidans ) , M . hungatei did not fully utilize all of the H2 and CO2/formate produced by S . fumaroxidans , even at high dilution rates ( Fig 4E and 4F ) . The net H2 production by the community indicates that S . fumaroxidans produces more H2 than M . hungatei’s needs and suggests the community can maintain higher M . hungatei to S . fumaroxidans ratios or that S . fumaroxidans could support faster growth of M . hungatei . At a dilution rate of 0 . 05 days-1 , S . fumaroxidans produces H2 in excess of M . hungatei’s energy needs until the relative M . hungatei to S . fumaroxidans biomass ratio reaches approximately 1 . 6:1 ( Fig 5 ) . These simulations suggest that invariant external H2 concentration requires high ratios of M . hungatei to S . fumaroxidans ( higher than has been proposed in the literature based on overall reaction stoichiometries ) , M . hungatei growing at faster rates than S . fumaroxidans ( e . g . , in batch culture ) , or a combination of the two . Furthermore , studies have shown that in coculture , S . fumaroxidans passes electrons to M . hungatei via formate , as well as H2 [38 , 39 , 56] . Coculture simulations predicted that formate could be exchanged in lieu of CO2 ( Fig 4B and 4C ) , without affecting the predicted bioreactor yields or species-weighted total flux ( pTMFA objectives ) . When formate is exchanged the formate dehydrogenases of S . fumaroxidans and M . hungatei facilitate the interconversion of formate to CO2 and H2 . Finally , steady-state metabolite concentrations in the coculture were predicted using thermodynamic variability analysis [33 , 34] ( S1 Text ) . Similar to a previous study of E . coli [34] , the majority of steady-state metabolite concentrations were not constrained by thermodynamics ( i . e . , the concentration ranges were the global concentration bounds of 0 . 01mM and 20 mM ) . However , hypotheses for some extracellular ( Table 2 ) and intracellular ( S5 Table in S1 Dataset ) metabolite concentrations could be made . The model predicts that propionate in the media must be greater than 0 . 36 mM for coculture growth to occur , and that acetate and CO2 concentrations must be less than 4mM and 0 . 65 mM , respectively . The model also predicts critical concentrations for H2 ( 0 . 0032 mM ) and formate ( 0 . 0020 mM , when formate is being exchanged ) , which must be maintained in order for methanogenesis to occur . All of these predictions were insensitive to both the dilution rate and biomass ratio ( M . hungatei: S . fumaroxidans ) .
The majority of the iMhu428 model content comes from the iMB745 reconstruction of M . acetivorans and still needs to be verified . Changes to model content could affect the conclusions drawn about M . hungatei behavior . Despite containing a complete methanogenesis pathway , the iMhu428 model was unable to identify the H+/Na+ transport stoichiometry of the energy-converting ( Eha- or Ehb-type ) hydrogenase ( EHA , 1 . 12 . 7 . 2 ) , which is thought to pump H+/Na+ while reducing ferredoxin [7] . The heterodisulfide reductase ( HDR , 1 . 8 . 98 . 1 ) can also reduce ferredoxin , and the iMhu428 model predicted HDR to be the only ferredoxin-reducing reaction required for methanogenesis . This observation is consistent with the observation that the expression of Eha/Ehb is considerably lower than that of HDR [57] . Additionally , different stoichiometries for other ion transport reactions important to methanogenesis remain thermodynamically possible . For example , the group contribution method predicted that tetrahydromethanopterin S-methyltransferase ( MTSPCMMT_CM5HBCMT , E . C . 2 . 1 . 1 . 86 ) could drive transport of up to 4 Na+ ions under standard conditions , instead of the 2 Na+ ions used in the iMhu428 reconstruction . Furthermore , some studies suggest the archaeal A1A0 ATP synthase is coupled to Na+ instead of H+ translocation [7 , 58] . When the iMhu428 model was modified to reflect this coupling , the model predicted the Na+/H+ antiporter was no longer active , as Na+ ions from MTSPCMMT_CM5HBCMT were directly used for ATP synthesis . Thus , while the modeled methanogenesis pathway is consistent with available data , it is not the only possibility . Analysis of the iSfu648 model revealed that H2 production is thermodynamically feasible in monoculture , implying there may be other biological reasons why H2 production is not normally observed under this condition , or that tighter estimates of thermodynamic parameters are needed . The iSfu648 thermodynamic model also has some important limitations . In particular , the iSfu648 model does not contain enough thermodynamic information to predict the directions of important electron transport reactions that involve ferredoxin , including the confurcating hydrogenase and formate dehydrogenase , the ferredoxin-oxidizing hydrogenase and formate dehydrogenase , and the RNF-type oxidoreductase ( S3 Table in S1 Dataset ) . This is because the group contribution method is unable to estimate the standard transformed Gibbs free energy of formation ( ΔfG'0 ) of ferredoxin , resulting in no ΔrG'0 estimates for these reactions . Fortunately , new quantum chemical approaches for estimating the thermodynamics of metabolism [59] may potentially provide additional ΔfG'0 estimates . This work also raises important questions about the appropriate mathematical basis for representing thermodynamic constraints . Previous studies used the ΔG'0 of groups directly when modeling thermodynamics [34] , and found that introducing uncertainty into a thermodynamic model of E . coli made the model computationally difficult to solve . In this work , using either the ΔG'0 of molecules or groups to model thermodynamics proved computationally difficult ( results not shown ) . Instead , only using ΔrG'0 as the basis for thermodynamic calculations enabled uncertainties in free energy estimates to be handled without any computational difficulties . However , using ΔrG'0 as a basis for thermodynamic calculations leads to larger uncertainties in ΔrG'0 and greater network flexibility , as it does not account for thermodynamic interconnectivity between reactions with shared metabolites . As a result , the model-predicted feasible ΔrG'0 range through a linear combination of reactions considerably exceeds the group-contribution predicted ΔrG'0 range of the overall reaction . In the future , thermodynamic interconnectivity should be captured using the ΔG'0 of molecules or groups , and optimization techniques are needed to improve the runtime performance of the resulting thermodynamic models . However , the use of ΔrG'0 as a basis for thermodynamic calculations is insufficient to explain why the iSfu648 model could still produce H2 under monoculture growth conditions . As described in S1 Text , ΔfG'0 was used as a basis for thermodynamic calculations when attempting to find additional constraints which would prevent H2 production . The failure to find such constraints indicates either that thermodynamics does not explain the absence of H2 production in monoculture , or that current thermodynamic models cannot capture this phenomenon . If it is the latter , more accurate group contribution methods with smaller error estimates may eventually be able to explain the role of thermodynamics in syntrophic associations . The thermodynamic coculture model of the syntrophic association between these species confirmed the role of formate and H2 in electron transfer in the community , and led us to hypothesize that total H2 consumption by the community indicates that M . hungatei cells are more abundant and/or faster growing than the S . fumaroxidans cells . The coculture model correctly predicted that both H2 and formate could shuttle electrons between members of this community . Formate may be preferred over H2 for electron transfer for thermodynamic reasons , as the ΔrG'0 of Eq 4 is more favorable ( less positive ) than that of Eq 3 in which formate is not exchanged . By exchanging formate in place of CO2 and H2 , S . fumaxoridans could sustain propionate oxidation at higher extracellular concentrations of formate than of H2 . Formate exchange could also be preferred due to differences in kinetics , diffusion , and/or volatility . These scenarios could stabilize the syntrophic association by enabling faster shuttling of electrons to M . hungatei . This work highlights some important obstacles to successful modeling of microbial consortia . In order for computational models of microbial communities to make meaningful predictions , individual species models must be integrated into a community model in a biologically relevant manner . Such integration will require an understanding of the objectives and constraints governing the behavior of each community member , and this work demonstrates that identifying the proper constraints and objectives requires extensive experimental characterization of the community . For example , neither maximization of growth rate nor maximization of ATP yield were sufficient for the iSfu648 model to predict the observed behavior of S . fumaroxidans . Additional constraints on reaction directions and flux ratios were required before iSfu648 could be combined with iMhu428 model to simulate the coculture . In addition , the iSfu648 model relied on data from gene expression and 13C NMR experiments , suggesting that constraint-based approaches will complement traditional top down ( ‘omics’ ) approaches [60] by enabling a mechanistic understanding of microbial interactions [24] .
The iMhu428 reconstruction of M . hungatei was built from the iMB745 reconstruction of M . acetivorans [41] . A preliminary draft reconstruction was built based on sequence homology ( using the RAVEN Toolbox [42] ) , but the reconstruction contained less than 200 genes ( S2 Dataset ) . Instead of performing extensive gapfilling , reactions from the iMB745 M . acetivorans reconstruction were copied into the M . hungatei reconstruction , with modifications to reflect key metabolic features of M . hungatei ( see S1 Text ) . Results from the RAVEN Toolbox and the KEGG SSDB [43] were used to map genes in M . acetivorans to M . hungatei and identify those reactions which have genomic evidence ( S2 Dataset ) . Finally , blocked reactions lacking genomic evidence were removed from the reconstruction . The final iMhu428 reconstruction contains 720 reactions , 428 genes ( associated with 493 reactions ) , and 639 metabolites . Of the 428 genes , 351 were added based on sequence homology , and 77 were added manually . The reconstruction is available in S2 Dataset and S3 Dataset in Excel and SBML formats . The iSfu648 reconstruction of S . fumaroxidans was built from KEGG [43] ( S6 Dataset and S7 Dataset ) using the RAVEN Toolbox [42] , which uses protein homology to identify the KEGG Orthology ( KO ) ID for each gene in a genome . The reactions and genes corresponding to that KO ID are then imported into the reconstruction . The resulting draft reconstruction was manually refined as described in S1 Text . The final iSfu648 reconstruction contains 874 reactions , 648 genes ( associated with 770 reactions ) , and 893 metabolites . The reconstruction is available in S4 Dataset and S5 Dataset , in Excel and SBML formats . Flux-balance analysis ( FBA ) [27] is a constraint-based technique for predicting the state of a metabolic network consistent with physiochemical principles . FBA identifies a flux distribution which maximizes cellular growth ( or some other objective function ) , subject to steady-state mass-balance and enzyme capacity constraints . Thermodynamics-Based Metabolic Flux Analysis ( TMFA , [33 , 34] ) extends FBA via the introduction of thermodynamic constraints , which require that the transformed Gibbs free energy of a reaction ( ΔrG' ) and its flux ( v ) have opposite signs . Estimates ( ΔrGest'0 ) and uncertainties ( SEΔrGest'0 ) of ΔrG'0 for the reactions in the reconstructions were obtained using a group contribution method [61] via the von Bertalanffy 2 . 0 Toolbox [62] . TMFA was implemented as previously described [34] , with additional details given in S1 Text . The mol files for metabolites in iMhu428 and iSfu648 are provided in S1 File and S2 File , respectively . pFBA [54] is a constraint-based approach which maximizes cellular growth while also minimizing total flux through the network ( a proxy for minimizing the total mass of enzymes required to sustain optimal growth through the network ) . pTMFA uses the same assumptions as pFBA while implementing the thermodynamic constraints of TMFA . pTMFA was implemented as a two-stage optimization process . In the first stage , growth rate is maximized via TMFA . In the second stage , the growth rate is fixed and the total flux through the network is minimized , subject to the same constraints as TMFA . Additional details on implementation can be found in S1 Text . For the coculture simulations , a community model of growth in a continuous stirred-tank reactor was developed that accounts for the biomass concentrations of each species . The model minimizes the species-weighted total flux through the metabolic networks subject to TMFA constraints for each species and mass balances around the entire reactor . Details on the specific implementation used in this work can be found in S1 Text . To avoid solving a mixed-integer non-linear program ( MINLP ) , the dilution rate and biomass concentrations for each species were fixed , resulting in a MIP . To explore the community behavior under a variety of operating conditions , the reactor dilution rate was systematically changed , while allowing unlimited propionate uptake by the reactor . The uptake fluxes for carbon and other nutrients used in the simulations are given in S1 Table in S1 Dataset . All simulations were performed using CPLEX 12 ( IBM , Armonk , NY ) accessed via the General Algebraic Modeling System , Version 23 . 9 . 5 ( GAMS , GAMS Development Corporation , Washington , DC ) . Estimates ( ΔrGest'0 ) and uncertainties ( SEΔrGest'0 ) of thermodynamic parameters were obtained using version 2 . 0 of von Bertalanffy and Matlab R2012b ( The MathWorks , Inc . , Natick , MA ) .
|
Natural and engineered microbial communities can contain up to hundreds of interacting microbes . These interactions may be positive , negative , or neutral , as well as obligate or facultative . Syntrophy is an obligate , positive interaction , in which one species lives off the metabolic by-products of another . Syntrophic associations play an important role in sustaining a variety of natural communities , including those involved in the breakdown and conversion of short-chain fatty acids ( e . g . , propionate ) to methane . In many syntrophic communities , electrons are transferred from one species to the other through small molecule diffusion . In this work , we expand the study of a two-member syntrophic , methanogenic community through the development and analysis of computational models for both species: the bacterium Syntrophobacter fumaroxidans and the methanogenic archaeon Methanospirillum hungatei . These models were used to analyze energy conservation mechanisms within each species , as well as small molecule exchange between the two organisms in coculture . The coculture model correctly predicted the exchange of both H2 and formate between the two species and suggested conditions under which these molecules would be fully metabolized within the community .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Thermodynamics and H2 Transfer in a Methanogenic, Syntrophic Community
|
The HIV-1 Vpu protein is expressed from a bi-cistronic message late in the viral life cycle . It functions during viral assembly to maximise infectious virus release by targeting CD4 for proteosomal degradation and counteracting the antiviral protein tetherin ( BST2/CD317 ) . Single genome analysis of vpu repertoires throughout infection in 14 individuals infected with HIV-1 clade B revealed extensive amino acid diversity of the Vpu protein . For the most part , this variation in Vpu increases over the course of infection and is associated with predicted epitopes of the individual's MHC class I haplotype , suggesting CD8+ T cell pressure is the major driver of Vpu sequence diversity within the host . Despite this variability , the Vpu functions of targeting CD4 and counteracting both physical virus restriction and NF-κB activation by tetherin are rigorously maintained throughout HIV-1 infection . Only a minority of circulating alleles bear lesions in either of these activities at any given time , suggesting functional Vpu mutants are heavily selected against even at later stages of infection . Comparison of Vpu proteins defective for one or several functions reveals novel determinants of CD4 downregulation , counteraction of tetherin restriction , and inhibition of NF-κB signalling . These data affirm the importance of Vpu functions for in vivo persistence of HIV-1 within infected individuals , not simply for transmission , and highlight its potential as a target for antiviral therapy .
The HIV-1 genes nef , vpu , vif and vpr are known as accessory genes and early in vitro studies showed them dispensable for viral replication in some tissue culture cell lines [1] . In vivo , however , these proteins are essential for the transmission and persistence of immunodeficiency viruses . Vpu , in particular , is thought to have been pivotal to the ability of HIV-1 group M to establish pandemic infection in humans following transmission from chimpanzees [2] , [3] . Expressed late in the viral life cycle , it functions during viral assembly to facilitate efficient egress of infectious viral particles , through the degradation of CD4 in the endoplasmic reticulum ( ER ) and the counteraction of the interferon-induced antiviral protein tetherin ( BST2/CD317 ) [4] , [5] . By antagonising tetherin , Vpu also acts to evade innate immune sensing of budding viral particles by repressing pro-inflammatory signalling events triggered by tetherin [6]–[8] . In recent years , Vpu has been implicated in other immunomodulatory functions , such as the downregulation of NTB-A/SLAMF6 [9] and poliovirus receptor ( PVR/CD155 ) [10] to evade NK cell recognition of HIV-1 infected cells , and the removal of CD1d from the surface of dendritic cells , inhibiting lipid antigen presentation to NK-T cells [11] . Furthermore , signature residues in the C-terminus of Vpu are associated with NK cell escape in KIR2DL2 positive individuals [12] . Vpu is found only in the SIVcpz/HIV-1 lineage of primate lentiviruses , yet its ability to counteract tetherin and downregulate CD4 is inconsistent throughout the members of this family . Vpu proteins from HIV-1 group M tested to date can perform both functions; the majority of available group N Vpu proteins weakly counteract tetherin and do not degrade CD4 , although show signs of adapting to human tetherin [13]; while in contrast , group O and P proteins can degrade CD4 but are fundamentally ineffective at counteracting tetherin [3] , [14]–[16] . The Vpu of the precursor virus , SIVcpz , can degrade CD4 but is ineffective against both chimpanzee and human tetherins; in infected chimpanzees , Nef performs this role by targeting a region of chimpanzee tetherin deleted in its human orthologue [3] , [17]–[19] . Vpu is absent from the genome of HIV-2 , therefore the envelope protein has adapted to the role of tetherin antagonist in these viruses [20] , whilst Nef alone downregulates CD4 . Thus , of all the immunodeficiency viruses able to infect humans , HIV-1 group M is the only virus group able to both degrade CD4 in the ER and counteract and ultimately degrade tetherin , suggesting that the Vpu protein may play a key role in the transmissibility and pathogenicity of this group , and potentially its pandemic status [3] . Most characterisation of group M Vpu thus far conducted has been of the prototypical molecular clone virus , NL4 . 3 ( reviewed in [4] ) , on panels of representative Vpus from different clades [3] , [14] , [15] , [21] , or on bulk-cloned proviral sequences [22] necessitating an in depth study of natural vpu alleles . In mice , tetherin activity moderates the replication and pathogenicity of murine retroviruses [23]–[25] , suggesting it plays an antiviral role in vivo . Overcoming the physical block to virus release is one obvious reason that Vpu tetherin antagonism might be essential for HIV-1 in vivo . However , whether tetherin can block cell-to-cell transmission of HIV-1 , likely to be the predominant mode of systemic viral spread in lymphoid tissue , is controversial and cell-type dependent [26]–[28] . The presence of tetherin at the virological synapse can , in some circumstances , enhance cell-to-cell virus spread [27] , in agreement with early observations that Vpu-deleted viruses spread faster in tissue culture [29] . Moreover , in all studies directly addressing the role of tetherin in cell-to-cell spread of HIV-1 the effects , either positive or negative , have been weak . Alternatively , downstream consequences of tetherin restriction in vivo , in particular the recently described pattern recognition activity of tetherin [6]–[8] , may put extra selective pressure on the maintenance of Vpu function throughout infection . Tetherin expression is upregulated on HIV-1 target cells in infected individuals [30] . Interestingly , sequence changes in Vpu have been documented in patients co-infected with hepatitis C virus after treatment with pegylated type-1 interferon [31] . This prompted us to question whether tetherin antagonism is important throughout HIV-1 infection in vivo , or whether functional variability in this attribute is tolerated after the virus has established a systemic infection . Using single genome amplification of vpu alleles from infected individuals and optimised assays for the three major functions of Vpu , we completed a comprehensive study of Vpu function in natural HIV-1 infection . Single genome amplification eliminates sample bias and PCR-based recombination and provides a representation of the proportion of viral alleles circulating at one time point , whilst allowing direct progression to tractable functional assays . In the latter feature , at present , it has an advantage over deep/next generation sequencing approaches . Furthermore , deriving vpu sequences from virions rather than cell-derived provirus is entirely representative of one particular timepoint , and less likely to contain defective variants in comparison . The aims of the study were twofold: to comprehensively characterise Vpu sequence variation , immune pressure and major functions from natural infection; and to inform current structure-function studies of Vpu by investigating naturally defective and sub-optimal Vpu proteins .
Vpu sequences were derived from actively replicating plasma virus from 14 HIV-1-infected individuals: 5 long-term non-progressors ( LTNP ) , 5 rapid progressors ( RP ) , and 4 normal progressors ( NP ) , detailed in Table 1 [32]–[34] . Patients were classified as follows , according to standard MACS criteria: individuals that progressed from seroconversion to AIDS in less than 5 years were designated RPs; 5–10 years NPs; and greater than 10 years for LTNPs . All individuals were treatment naïve , both during and prior to the time of sampling . Up to three different time points were obtained from each individual , ranging from seroconversion ( 0 years ) to 10 . 4 years , with 1–2 year and 3–4 year time points obtained for each individual where possible . To represent fully the vpu repertoire in each plasma sample and to maximise the probability of isolating representative minor viral variants , at least 29 sequences were obtained from each sample , yielding a total of 851 vpu sequences from all 25 plasma samples ( Table 1 ) . All 851 nucleotide sequences obtained were aligned and assembled into a maximum likelihood phylogenetic tree ( Figure 1 ) , with vpu sequences from each infected individual forming a monophyletic group , in accordance with their distinct origin . Sequences from individuals with different progression rates to disease did not cluster in proximity to each other , indicating the lack a direct relationship between a specific vpu sequence and pathogenic outcome . Individual phylogenetic trees of vpu sequences from each of the 14 individuals are shown in Figure S1 . The mean intra-patient nucleotide and amino acid diversity for each individual vpu repertoire is shown in Table 1 . vpu sequence diversity did not correlate with disease progression rates , with individuals within the groups harbouring a range of sequence diversity levels ( LTNP 1 and 4 , for example ) . As expected , mean intra-patient nucleotide and amino acid diversity increased over time in all individuals . There was no correlation between genetic distance and the number of sequences obtained per sample , suggesting that the viral repertoire in each sample had been fully represented ( Figure S2 ) . Vpus from LTNP 1 showed the highest level of genetic diversity , as evident from Figure 1 and Table 1 , however , it should be noted that an extra 10 . 4 year time point was analysed from this individual; sequence diversity was comparable to others from the same progression group at equivalent time points . We found no indication of APOBEC3-mediated changes acting on the individual vpu populations ( data not shown ) . Of the 851 vpu sequences obtained , 456 had unique nucleotide sequences , and 304 unique amino acid sequences . Of these 304 alleles , five contained readily detectable mutations ( i . e . deletions or frame-shifts ) , specifically: two contained a premature stop codon ( LTNP2v14_4_51 and LTNP5v22_5_71 , resulting in a 6- and 1-amino acid C-terminal truncation , respectively ) , one contained a frame-shift ( RP2v16_1_5 ) , and two contained a 1-amino acid N-terminal deletion ( LTNP1v11_4_3 and 5_38 ) . The other 299 Vpus were 81 amino acids in length and thus potentially functional . Since a single amino acid change can impact on the function of a protein , all 304 Vpu alleles were cloned and tested in standard functional assays for CD4 downregulation and tetherin counteraction . Samples were weighted according to how many genomes were isolated with a particular amino acid sequence , thus representing the proportions of functional and non-functional Vpus present in a given sample ( Figure 2 ) . Vpu from the HIV-1 clade B molecular clone NL4 . 3 was used as the prototypical Vpu in all assays , to which the functions of patient-derived Vpu proteins were compared . Mutant Vpus derived from NL4 . 3 with defects in tetherin counteraction ( A14L ) , or both tetherin counteraction and CD4 downregulation ( S52 , 56A and AW14 , 22LA ) were included in each assay as negative controls . We also tested Vpus derived from a panel of HIV-1 clade B transmitted/founder viruses as representatives of earliest available replicating virus [35] . Assay cut-offs were determined by the performance of the entire Vpu population , with the threshold for sub-optimal or defective activity set at one standard deviation below the mean: for CD4 downregulation this was 73 . 7 ( mean Vpu function 90 . 3% , standard deviation 16 . 6 ) ; for tetherin counteraction this was 81 ( mean Vpu function 114% , standard deviation 33 ) . Based on these criteria , 17 Vpus were suboptimal/defective for CD4 downregulation , and 41 for tetherin counteraction . Interestingly , the founder virus Vpus displayed a spread of function representative of the 304 patient-derived alleles , with CH040 Vpu showing sub-optimal activity . There were no discernible correlations between either of the two Vpu functions and disease progression groups or time post-seroconversion . Neither could we detect a correlation between tetherin counteractivity and viral load; although in six of the eight individuals with more than one time point , anti-tetherin function did increase with an increase in viral load ( data not shown ) , but changes were not significant . CD4 downregulation activity was highly maintained across all individuals and time points . Vpus from the same individual had a narrow range of function , with defective Vpus set apart from the rest of the group , indicative of an intrinsic activity of each Vpu population . The spread of tetherin counteraction was broader than that of CD4 downregulation , perhaps due to a more complex mechanism and more regions of the protein involved in tetherin downregulation and degradation . In individuals in which there was a discernable group of suboptimal Vpus , the group was diminished in number at the later time point ( NP 2 , NP 3 , LTNP 3 ) , suggesting ongoing pressure for Vpu to maintain optimal function throughout infection . To assess the impact of single , and thus potentially transitory , variants , the functional data from Figure 2 were re-plotted after the removal of all Vpus represented by one single genome , showing only the variants represented by multiple genomes ( Figure S3 ) . For the most part , the data remain unaffected , with only the degree of significance changing over time in some individuals . Specifically , for CD4 downregulation the decreases seen for NP 1 and LTNP 5 had higher p values when single variants were removed , and the NP 2 increase became significant ( p = 0 . 042 ) . For tetherin counteraction , the increase seen for NP 2 had a higher p value , as did the decreases seen for LTNP 1 and LTNP 5 . For some individuals , removal of single variants lead to all values for one time point being identical , and in these cases statistical analyses could not be performed ( LTNP 3 , LTNP 4 ) . The majority of our natural Vpu proteins had tetherin counter-activity superior to that of the NL4 . 3 Vpu prototype . The ability of NL4 . 3 to down-regulate CD4 , however , appeared near-optimal compared to primary Vpu proteins . Direct comparisons between NL4 . 3 Vpu and three representative patient-derived Vpus from each of the three progression groups ( RP2v16_2_87 , NP2v11_2_1 , and LTNP1v4_1_67 ) , confirmed that NL4 . 3 performed notably poorer than typical clade B Vpus in tetherin counteraction ( Figure 3 ) . The three natural Vpus differed in sequence by 7 to 10 amino acids from the sequence of the Consensus B Vpu obtained from the Los Alamos National Laboratory HIV database . The expression of the patient-derived Vpus was not greater than that of NL4 . 3 by Western blot ( Figure 3 ) , and in titration experiments , up to 100% more infectious virus is released in the presence of the patient-derived Vpu compared to NL4 . 3 Vpu . The ability to downregulate CD4 was optimal for the four proteins ( RP2v16_2_87 bearing 101% activity relative to NL4 . 3 , NP2v11_2_1 89% , and LTNP1v4_1_67 86% ) , supporting the notion that NL4 . 3 is inferior to natural Vpu proteins only in tetherin counteraction . The superiority of the patient-derived Vpus was further demonstrated by their suppression of tetherin-mediated NF-κB activation . In transient tetherin signalling assays , in which tetherin is overexpressed to mimic receptor clustering and activate NF-κB [7] , various Vpu constructs were titrated and assessed for their ability to reduce activation of an NF-κB reporter construct by tetherin . At 25 ng of Vpu the residual tetherin signalling activity in the presence of RP2v16_2_87 , NP2v11_2_1 and LTNP1v4_1_67 was 34 , 30 and 30% respectively , compared to 63% in the presence of NL4 . 3 . All patient-derived Vpus almost completely abolished tetherin signalling at the higher Vpu expression level of 100 ng . Comparison of the NL4 . 3 and patient-derived Vpu amino acid sequences highlights differences in the C-terminal portion of the cytoplasmic tail , notably in the 2nd alpha-helix where putative trafficking domains and acidic patches are positioned differently relative to the conserved phosphorylated serines ( Figure 3 ) . When compared for CD4 downregulation and tetherin counteraction activity , each Vpu had a unique functional profile , as shown in Figure 4 . The vast majority of Vpus were able to perform both functions ( n = 263; 86 . 5% ) , yet there were sufficient numbers of defective proteins to merit investigation of structure-function relationships . Vpus with levels of activity ranging from defective to sub-optimal ( defined as 0–81% of NL43 activity for virus release; 0–73 . 7% for CD4 downregulation ) were categorised according to whether they had defects in tetherin counteraction only ( n = 23; 7 . 6% of all Vpus ) , CD4 downregulation only ( n = 7; 2 . 3% ) , or both ( n = 11; 3 . 6% ) . Of note , there were more Vpus defective for tetherin counteraction alone than there were for CD4 downregulation only , and the overall range of function for tetherin counteraction was broader than that of CD4 downregulation . Comparing Vpu sequences from different viral isolates in order to identify regions of functional interest can often be problematic due to multiple differences between given sequences [14] , [21] . However , the advantage of using sequences obtained by SGA is that , in the majority of cases , each defective or suboptimal Vpu has a functional relative that differs by only one or two amino acids . Thus , by comparing the sequences of defective Vpus to their closest functional relatives , and then to the entire Vpu repertoire , in most cases the amino acid change or changes responsible for the defect could be identified ( Figure 4 , Figure S4 and Table S1 ) . Proteins defective for both CD4 downregulation and tetherin counteraction ( i . e . less than the 81% cutoff for tetherin and 73 . 7% for CD4 ) contained a frame-shift ( n = 1 ) , an A19E change in the transmembrane domain ( n = 1 ) , mutations of the highly conserved regions just prior to ( E29K; n = 2 ) and within ( II43 , 46SL , R49G , R49T; n = 3 ) the first alpha-helix , and in the DSGNES hinge region between the two cytoplasmic alpha helices ( D52V , SN53 , 55RH , S53N , E58K; n = 4 ) , which contains two phosphorylated serines essential for interactions with the E3 ubiquitin ligase complex SCFβ-TrCP , central to Vpu's function . Since the expression of Vpus with defects in both functions could not be assumed , expression of these proteins was verified by Western blot analysis , and although variable , all but one Vpu could be detected . The latter , when compared with known functional proteins from the same sample , was from a population of Vpus not recognised by the anti-Vpu antibody used for Western blot analysis ( Figure S4 ) . Since there were only four Vpus with a defect in CD4 downregulation alone , this presented fewer opportunities for determining regions specific only to this function . Indeed , in contrast to tetherin counteraction , there is little consensus in the literature regarding individual amino acids or motifs in Vpu specifically governing CD4 downregulation . However , of the three in which specific changes could be assigned to loss of function , these mapped to conserved residues in the first alpha helix of the Vpu cytoplasmic tail ( n = 1 ) and transmembrane region ( n = 2 ) , specifically I17T , V22A and I39L . One caveat to the CD4-only defects is that , while the tetherin functions for all of them were more than 81% that of NL4 . 3 , in many cases they were still suboptimal compared to the better performing Vpus in the data set . Of note , the transmembrane residues assigned to CD4 downregulation defects were highly conserved . Tetherin-specific functional mutations were tracked to the transmembrane domain ( n = 14 ) , to conserved residues in the first alpha-helix ( E48 , n = 2 ) , to the conserved DSGNES hinge region ( n = 3 ) ; to the ExxxLV motif ( and flanking residues ) in the second alpha-helix ( n = 3 ) ; and to a conserved tryptophan in the Vpu C terminus ( n = 3 ) ; with 2 Vpus with unassignable defects ( Figure 4 , Figure S4 ) . At least two regions in Vpu have previously been assigned specific functions in the context of tetherin counteraction: in the transmembrane domain , alanines at position 15 , 19 and a tryptophan at position 23 ( positions 14 , 18 and 22 in NL4 . 3 for reference ) , aligned along one face of the transmembrane helix , form an interacting surface with the tetherin transmembrane region [36]; in the second alpha helix of the cytoplasmic domain , an ExxLV motif , a putative sorting signal , plays a role in trafficking and degradation of Vpu/tetherin complexes [37] . A high proportion of the mutations that affected only tetherin counteraction mapped to an A15 change to a valine or threonine , ( n = 14 ) , and for the most part resulted in a modest reduction in tetherin counteraction . Since 12 . 2% ( n = 37 ) of all Vpus contained a valine at this position , and a further 1 . 6% ( n = 5 ) a threonine , and not always immediately conferring a disadvantage in comparison with NL4 . 3 , the effect at this position is clearly dependent on context and may potentially weaken the interaction with tetherin . However , when comparing V15 and T15 Vpus with matched Vpus from the same infected individual with alanines at this position ( when available ) , rather than with NL4 . 3 Vpu , all demonstrated at least a 50% relative defect in virus release ( Table S1 ) . Interestingly , in two individuals , NP 2 and LTNP 3 , V15 or T15 Vpus make up a large population of sub-optimal Vpus ( 35 . 3 and 100% of the 1–2 year time point respectively ) , with the overall function of these time points falling at or below the level of NL4 . 3 . In both cases these are significantly fewer in proportion in the following time point ( using Fisher's exact test: NP 2 p = 0 . 043; LTNP 3 p = 0 . 0046 ) , indicative of them being selected against , and the overall function of the subsequent time point is significantly higher ( Figure 2; NP 2 , LTNP 3 ) . In contrast to the variation seen at position 15 , only one Vpu contained a mutation at position 19 ( NP1v5_1_80 , A19E ) , leading to a loss of anti-tetherin function and a severe ( 2-fold ) defect in CD4 downregulation , whereas W23 was completely conserved , highlighting its critical role in both major functions of Vpu [36] , [38] . Two other mutations in the transmembrane domain led to a specific loss of anti-tetherin function: I9M and I16E . Whilst not forming part of the “alanine face” of Vpu , these polar or charged residues are adjacent and may impact upon accessibility of the tetherin binding interface . Interestingly , mutations occurring at the DSGNES β-TrCP binding site that occurred between the two phosphorylated serines , N55H ( n = 2 ) and E56G ( n = 1 ) , were highly specific and deleterious for anti-tetherin activity , but were functional for CD4 downregulation . Mutations at or outside these phosphorylated serines , as described earlier , had severe effects on both functions and behaved essentially as the S52 , 56A mutant used as a functionally defective control in the function assays . Since β-TrCP is essential for CD4 downregulation by Vpu [39] , these data suggest that there is a separable element to the function of this region that is independent of SCF E3 ubiquitin ligase recruitment . As we had thoroughly examined 304 Vpus for CD4 downregulation , and tetherin counteraction , we also decided to test those defective for counteraction of both tetherin functions for their ability to downregulate cell-surface tetherin expression ( Figure S5 ) . We found the majority of Vpus defective for virus release maintained the ability to downregulate tetherin , possibly due to the majority of the mutations tested mapping to the DSGNES , which has residual function for tetherin downregulation [36] , [40] , and to the second cytoplasmic helix , previously suggested to have intermediate impact on internal tetherin sequestration [41] . The dichotomy of function illustrated in Figure 4 and Figure S4 , wherein observed mutations of the D52 , S53 , and E58 lead to a severe defect in both tetherin antagonism and CD4 downregulation , whilst mutations of N55 and E56 disproportionately affected tetherin counteraction , warranted further investigation . The β-TrCP binding consensus sequence is DpSGxxpSE , where both serines are phosphorylated , and in all β-TrCP substrates other than Vpu ( e . g . IκBα , β-Catenin , CDC25B ) , the amino acid adjacent to the glycine is hydrophobic , packing into a hydrophobic patch in the binding groove of β-TrCP . In Vpu , however , this residue is a highly conserved hydrophilic asparagine , mutation of which to histidine results in a dramatic reduction in the ability of the Vpu to counteract tetherin and promote virus release . We therefore set out to determine whether this functional defect was due to a reduced or abolished ability of the Vpu to bind β-TrCP by performing Vpu and β-TrCP co-immunoprecipitions . Using the closest functional relative from the same infected individual as a positive control , we compared the binding of all Vpus that had mutations in the DSGNES region . As expected , we observed no β-TrCP binding by Vpus containing mutations of S53 , D52 , and E58K . However , we observed a robust binding of β-TrCP by the N55H and E56G natural Vpu mutants ( Figure 5 ) , suggesting an alternative tetherin-specific defect imposed by these mutations . Given the demonstrated superiority of a select few patient-derived Vpus to suppress tetherin-mediated NF-κB activation ( Figure 3 ) , we next tested all 304 Vpus in order to obtain both a full picture of signalling suppression in natural Vpu proteins , and also to discern potential residues in Vpu specifically involved in this function that have hitherto been uncharacterised . To date , there have been no reports of regions of Vpu required to specifically suppress tetherin signalling , although a generalised suppression of NF-κB activation upon over-expression of NL4 . 3 Vpu has been linked to the conserved β-TrCP binding site [42] . As with tetherin antagonism , there was a broad range of signalling-suppressive function , with some time points containing clusters of inferior Vpu function ( Figure 6 ) . Interestingly , in several individuals , including those from whom a seroconversion sample was available , signalling suppression was higher in the early time point and significantly declined over time ( Figure 6; NP2 , NP3 , LTNP3 , LTNP4 ) . In these individuals tetherin antagonism for virus release increased over time ( Figure 2 ) , indicating a trade off between the two elements of tetherin counteraction , and that the ability of Vpu to suppress tetherin-mediated signalling is not wholly determined by the physical counteraction of tetherin . Since the Vpu profiles of 14 infected individuals were not similar when compared for their ability to antagonise tetherin to promote virus or to suppress signalling , we investigated whether differences in these two activities could be assigned to specific amino acid changes not critical for the promotion of virus release . Taking the same approach as that used to compare tetherin antagonism and CD4 downregulation , functional profiles of all 304 Vpus were plotted ( Figure 7 ) . As is evident from Figure 6 , we observed no correlation between the ability of Vpu to physically antagonise tetherin and its ability to suppress tetherin signalling . Mutations that affected both functions were found in the DSGNES motif , a frameshift , the highly conserved R49 and E51 in the first alpha helix and A19 in the transmembrane domain . Interestingly , there were a considerable proportion of Vpus that were still able to counteract tetherin for virus release , but had defects in signalling suppression . These mapped to three conserved residues: G59 and E62 in the second alpha helix of the cytoplasmic domain , and R45 in the first alpha helix . A cluster of Vpus had defects that mapped to A50V or -T changes , which accounted for the suboptimal activity of the majority of Vpus isolated from NP1 ( Figure 6 ) . One more tetherin signalling-specific defect , I17T , was also defective for CD4 downregulation; all others were only defective for this particular function ( see Table S1 ) . One Vpu with a major defect in tetherin antagonism , II43 , 46SL , was still able to reduce NF-κB activation , and more modest mutants such as E29K , which were also defective for CD4 downregulation , maintained the ability to reduce tetherin signalling . Minor but common tetherin antagonism defects , A15V or –T , had no impact on the ability of these Vpus to suppress tetherin signalling . The ability of many Vpus to suppress tetherin signalling independently of their ability to promote virus release prompted us to investigate whether Vpu possessed a global NF-κB suppression activity , mediated through the sequestration of β-TrCP , as previously reported [42] , [43] . To test this we looked at the effect of increasing concentrations of various Vpu proteins on the activation of NF-κB by MAVS , a central adaptor protein in NF-κB activation pathways triggered by RIG-I-like RNA sensing receptors . First we compared the ability of a highly active patient-derived Vpu ( RP2v16_2_87 ) to counteract tetherin- and MAVS-mediated NF-κB activation , along with NL4 . 3 and known mutants thereof ( Figure 8 ) . RP2v16_2_87 Vpu was highly effective in suppressing NF-κB activation by both tetherin and MAVS , with an 88% and 94% reduction in signalling by both molecules , respectively , at the highest concentration tested ( Figure 8A ) . NL4 . 3 , in contrast , showed a weaker but dose-dependent ability to suppress tetherin signalling , but was severely defective for the inhibition of MAVS signalling , with an effect seen only at the highest concentration of 100 ng . The S52 , 56A NL4 . 3 mutant , unable to interact with β-TrCP , had no signalling-suppressive activity against either tetherin or MAVS , whereas the A14L tetherin binding mutant was able to partially inhibit NF-κB activation by both proteins at higher concentrations , consistent with the notion that Vpu mediates a concentration-dependent generalised inhibition of NF-κB activation that is independent of its ability to physically counteract tetherin . We next examined patient-derived Vpus that showed differential ability to counteract tetherin to promote virus release and to suppress signalling . Of these , an E29K mutant that was defective for both tetherin counteraction and CD4 downregulation was highly active in suppressing both tetherin and MAVS-mediated NF-κB activation . Conversely , G59R , E62G and A19E mutants were all impaired , to various degrees , for their ability to suppress both tetherin- and MAVS-mediated activation of NF-κB . To investigate whether certain amino acid changes were selected for within a given viral pool , either due to immune escape or functional advantage , we performed positive selection analyses on the complete vpu sequence sets from each infected individual . Overall , the vpu gene was found to be under purifying selection ( mean dN/dS ranging from 0 . 20 to 0 . 72 across individuals ) , with the identification of several individual amino acids under positive selection pressure ( Figure S6 ) . For the purposes of the population-level positive selection analysis , only the part of vpu that does not overlap with the env reading frame ( codons 1–54 ) was included in the analyses . For the separate patient analyses , positions that were found to be under positive selection that fell in the overlap were assessed on an individual basis ( for details see Materials and Methods ) . Few of the codon positions under positive selection were common to more than one individual , although positions in the N-terminus and transmembrane domain of the protein were frequently selected for ( see Table 2 , LTNP 1 , 3 , 5 , NP 1 , 2 , RP 1 and 2 ) . We found no positively selected sites associated with patterns of disease progression . Alignments of the amino acid sequences from each plasma sample show a regional clustering of mutations indicative of immune pressure , with positions undergoing positive selection often falling within these areas . We speculated that the regions of concentrated variation might coincide with T cell epitopes , previously poorly characterised specifically for Vpu , and that immune escape was principally driving the variation in the vpu gene . To ascertain the CD8+ T cell epitope potential of the Vpu sequences , the majority Vpu sequence from each time point was entered into a T cell epitope prediction algorithm ( IEDB MHC Class I prediction method version 2009-09-01B ) , tailored to the Class I HLA haplotype of the corresponding infected individual ( Table S2 ) . Overlaying the predicted CD8 T cell epitopes with amino acid sequence alignments demonstrates an accumulation of mutations in regions putatively targeted for presentation to CTLs , and often overlapping with sites under positive selection ( Table 2 ) . The co-localisation of predicted epitopes and positively selected amino acids explains the apparently random location of such residues . Furthermore , ordering individuals by genetic distance ( mean nucleotide substitutions/site at time point 1–2 years ) illustrates that those with the most variable Vpu repertoires also have the highest number of predicted CD8 T cell epitopes , allowing us to speculate that it is CD8 T cell pressure driving vpu variation , and that positive selection acting on apparently random positions is an indication of ongoing diversification within and around putative epitopes . Interestingly , in LTNP 1 , 3 and 5 , the individuals with the highest number of predicted T cell epitopes , a significant drop in one or both functions can be seen over time . Of note , one of the positions undergoing positive selection in LTNP 5 was residue 73 , a position previously linked to NK cell escape in KIR2DL2 individuals [12] ( Table 2 , Tables S1 and S2 ) . Upon further investigation we observed at least one change at this position , or at the associated position 70 , in all but two KIR2DL2 positive individuals ( LTNP1 , 2 , 3 and 5 , NP 1 and RP 2 ) ; in contrast , these residues were invariant in all KIR2DL2 negative individuals ( LTNP 4 , NP 2 and 3 , RP 3 , 4 and 5 ) . To investigate further the association between predicted T cell epitopes , immune escape and Vpu variation in more detail , we selected the vpu repertoire with the highest genetic diversity ( LTNP 1 ) , and compared function and sequence changes over time with predicted T cell epitopes and sites undergoing transient or pervasive selection ( Figure 9 ) . Despite cumulative mutations occurring in 20% of the protein ( 16 of 80 amino acids , excluding start and stop codons ) , pervasive or episodic selection acting at five positions ( i . e . codon positions 2 , 4 , 7 , 9 and 16 ) , and predicted high affinity T cell epitopes spanning the bulk of its length , every Vpu isolated from this individual was deemed functional by our classification .
Using single genome sequencing we have carried out a full characterisation of the sequence and function of the HIV-1 vpu gene throughout infection , and demonstrate that the Vpu protein has a considerable capacity for diversification and adaptation , consistent with it being one of the most variable regions of the HIV-1 genome [44] . In the face of predicted CD8 T cell pressure and significant sequence variation , it is able to maintain function regardless of disease stage or severity , with no indication of hierarchy of function . Vpu function is strictly maintained throughout infection , as shown by fully functional Vpus obtained from transmitted/founder viruses , from seroconversion time points and from viruses isolated more than 10 years post-infection . All three functions tested – CD4 downregulation , tetherin counteraction for virus release , and inhibition of tetherin-mediated NF-κB activation – were maintained , with the vast majority of proteins ( 96 . 7% ) active in at least one function . Of the ten Vpu proteins defective for all three functions , none were found in subsequent time points , suggesting that seriously defective variants do not persist over time . More minor defects in a single function did persist over time in certain individuals , for example those impaired for signalling suppression seen in NP 1 , but the real impact of modest defects in vivo is difficult to gauge . It is also possible that , for suppression of signalling , strict maintenance of function is more important at early stages of infection , and declines with time , as seen most notably in NP 3 . For tetherin counteraction , maintenance of function reflects other reports of immunodeficiency viruses responding to the pressure exerted by tetherin , including the recent characterisation of a HIV-1 group N Vpu that has evolved to become an efficient tetherin antagonist [13] , the demonstration of acquisition of tetherin antagonism in the Env proteins of nef-deleted simian immunodeficiency viruses [45] , and the reacquisition of tetherin counteractivity in Nef following experimental infection of chimpanzees with HIV-1 [46] . Furthermore , studies of HIV-1/hepatitis C-co-infected individuals have demonstrated that , following treatment with pegylated interferon , an increased expression of tetherin in peripheral CD4+ T cells correlates with a significant reduction of HIV-1 viral load , with some indication of compensatory mutations in Vpu [31] . Selective pressure exerted by tetherin is indicative of its multiple antiviral effects: not only its ability to physically prevent the release and spread of virus particles , its role as a pattern-recognition receptor and potential enhancer of antigen presentation [47] , but also the potential for enhanced antibody/complement opsonisation and NK cell recognition that may be downstream consequences of virion restriction . This is manifest in the observation that populations of sub-optimal Vpus with specific defects in tetherin counteraction in early time points , such as the group of Vpus with A15V in NP2 and LTNP 3 , are found in significantly lower frequencies at the subsequent time point , indicative of selection against Vpus with inferior tetherin binding activity . Our previous work demonstrates the ability of tetherin to induce an NF-κB-mediated proinflammatory signal [7] , and here we thoroughly examine the ability of 304 primary Vpus to counteract tetherin signalling . The suppression of tetherin-mediated NF-κB activation was observed at a high level across the patient groups particularly at early time points , with the notable exception of NP1 . However , while the majority of Vpus were superior to NL4 . 3 Vpu in both functions , there was no direct relationship between the ability to promote virus release and the ability to suppress signalling; in some individuals an increase in the former function over time was mirrored by a decrease in the latter . This prompted us to investigate whether signalling suppression , particularly by those Vpus with defects in direct tetherin antagonism , was in part due to a previously recognised [42] , [43] , and more recently expanded [48] , intrinsic ability of Vpu to globally suppress NF-κB activation . This ability of Vpu is primarily driven by its binding of β-TrCP , a component of the SCF E3 ubiquitin ligase complex that is required for the ubiquitination and degradation of IκB , and subverted by Vpu for the degradation of its target proteins [49] . Indeed , we confirmed that NL4 . 3 Vpu was able to suppress NF-κB activation by both tetherin and MAVS , but only when overexpressed , i . e . at levels unlikely to be found in an infected cell . Our patient-derived Vpus , however , were able to reduce NF-κB activation even at lower levels of expression , with a complete ablation of signalling occurring at higher Vpu concentrations , suggesting that this may indeed be an important role of Vpu in vivo . Furthermore , the observation of a significant decline in signal-suppressive function over time in several individuals , in contrast to the other two functions examined , as well as the high activity observed in founder virus-derived Vpus , may be indiciative of this activity being most important in early stages of infection . In addition to fully characterising natural vpu alleles , the secondary aim of this study was to identify determinants of the protein that are required for one or all functions . To the previously precisely defined regions involved in tetherin antagonism ( A15 , A19 , W23; E63 , L67 , V68 [4] , [36] , [37] ) , we contribute I9 , A16 , E29 , II43 , 46 , E48 , R49 , E51 , N55 , E56 and W76 . At positions recently highlighted by McNatt and colleagues to interact with tetherin ( I5 , A8 , V21 , V22 , V26 , I27 , I28 ) , we see variation amongst our patient-derived Vpus and encountered no changes here that impacted on tetherin counteractivity . However , an I9M mutation that was attributed to a serious defect in tetherin antagonism ( Figure 4 ) , is adjacent to a residue proposed to interact with tetherin [41] . V21 and V22 residues , also indicated as interacting residues [41] , we found had more influence on CD4 downregulation and little impact on tetherin counteractivity . The N55H and E56G mutations are particularly interesting , since they occur within the DSGNES motif containing the phosphorylated serines that mediate interaction with β-TrCP , and yet these particular changes leave CD4 downregulation largely intact . While previously suggested to impact on tetherin antagonism [50] , the lesser impact on CD4 downregulation promted us to investigate this further . We were able to show that Vpus with N55H or E56G mutations are still able to bind β-TrCP . Since other mutations in the DSGxxS diserine motif ( e . g . D52V , S53N , E58K ) had severe effects on β-TrCP binding and CD4 downregulation , this suggests a dual function of this region in accordance with previous demonstrations that β-TrCP is not strictly required for tetherin trafficking by Vpu [51] . It is possible that this reflects an as yet unidentified Vpu co-factor , or involves facilitating access to either of the two cytoplasmic alpha helices . In this respect , it is interesting to note that acidic-dileucine motifs , such as the ExxxLV motif of Vpu , have previously been associated with phosphorylation in the trafficking of the cation-independent mannose-6-phosphate receptor [52] . Examination of the ability of 304 different Vpu proteins to suppress tetherin-mediated NF-κB activation revealed a number of previously uncharacterised residues important for this function . Residues that were important for inhibition of tetherin signalling , but not for the other two tested functions , mapped to G59 and E62 in the second alpha helix , and A50 in the first alpha helix . Vpus containing G59 and E62 were likewise partially defective in inhibiting NF-κB activation by MAVS when tested over a range of concentrations , therefore indicating that this little characterised activity of Vpu involves residues in Vpu beyond that of the DSGNES β-TrCP binding site . Conversely , Vpu proteins that were defective for tetherin counteraction ( e . g . E29K ) , maintained the ability to suppress tetherin signalling , and NF-κB activation in general , presumably through possessing an intact β-TrCP binding site . The fact that this ability of Vpu involves regions of the protein beyond the β-TrCP binding site , may indicate that the mechanism of signal suppression is more complex than the sequestration of β-TrCP . To the more elusive residues involved in CD4 downregulation , including V21 , S/T24 [38] and L67 [53] , we add I17 , V22 and I39 , in addition to the E29 , I43 , I46 and R49 residues also found to affect tetherin counteraction . We found no mutations in the second alpha helix that may be attributed to CD4 interaction , as previously suggested [54] , although this may be due to more conservation in this area and therefore a lack of mutants with potential functional defects . V22 ( V21 in NL4 . 3 Vpu ) has previously been reported to have a mild effect on CD4 downregulation , as we show here [38] . I17 and I39 are highly conserved residues , and to our knowledge have not previously been implicated in CD4 downregulation , although they do fall within the transmembrane domain and first alpha-helix , two regions other than the DSGNES β-TrCP binding region previously reported to be important for CD4 downregulation [4] , [55] . It is perhaps surprising , considering that few other members of the immunodeficiency viruses have this capability , that the CD4 degradation activity of virtually all Vpus tested is so strictly maintained . The reasons for this are unclear; all known immunodeficiency viruses possess an activity in Nef that induces the endocytosis of CD4 from the infected cell surface , with only the HIV-1 groups M , O and P , and the SIVcpz viral lineage employing a further Vpu-induced CD4 degradation step in the ER . Yet , there is little doubt from our data that the degradation of CD4 in the ER is absolutely required by HIV-1 in vivo , and there is no suggestion that any redundancy of function exists between Vpu and Nef , or that a reduction in this function is tolerated over time . HIV-1 envelope affinity for CD4 is reportedly higher than that of tested SIV envelope proteins , thus it has been proposed that Vpu is required to effectively chaperone the Env protein through the ER , thus avoiding this high-affinity interaction and subsequent loss of Env integrity [56] . The importance of other recently reported functions of Vpu remains to be explored . Vpu plays a further role in the modulation of immune recognition of the infected cell through downregulating the NK cell activating ligand NTB-A [9] , and through reducing the surface expression of PVR and CD1-d [10] , [11] . Thus far , studies comparing the effects of Vpu in vitro and in humanised mice have demonstrated a clear effect of Vpu on CD4 and tetherin , with modest effects on NTB-A and CD1d [13] , [57] , [58] . The mechanisms also appear to differ , with the serines central to CD4 downregulation and tetherin counteraction not required for downregulation of cell surface NTB-A [9] . It will be interesting to ascertain whether the minority of non-functional alleles isolated in this study have residual activity against either of the recently characterised targets , and whether they are present in circulating virus strains because they modulate the recognition of the infected cell by NK or NK-T cells . We also see evidence of Vpu's immunomodulatory function in signature residues at its C terminus ( amino acids 70 and 73 ) previously linked to NK cell escape . Indeed , in KIR2DL2 positive individuals we detect ongoing variation at these positions; interestingly , this was most apparent in long-term non-progressors , and in one such individual we observe positive selection acting at position 73 , in accordance with the amino acid position associated with NK cell escape characterised by Alter et al [12] . Predicted CD8 T cell pressure coincides with positions of the protein we detect as being under positive selection . In at least one individual ( LTNP 5 ) we see evidence of escape from a high affinity T cell epitope at the later time point . In others , mutations occur in flanking residues of the peptide , potentially affecting peptide processing . In the two individuals with the highest variability and highest predicted T cell pressure ( LTNP 1 and 5 ) , we see a significant reduction of overall tetherin antagonism over time; however , as discussed above , the levels do not drop below that of NL4 . 3 and thus we predict would attain the threshold of activity required to manage tetherin in vivo . The demonstration of continuous pressure on the virus to maintain high levels of Vpu function , and our detailed analysis of Vpu sequence-function relationship , puts forth strong support for the development of antiretroviral compounds targeting Vpu , whilst providing an excellent resource for the future study of disease-relevant Vpu alleles . In particular , we provide data on the regions of Vpu common to two or more functions , and those that appear to be specific to one . We demonstrate that gross defects are not tolerated , making Vpu a potential target for drug development; yet , we stress the importance of assessing multiple parameters of accessory gene function . In this respect , replication in culture assays may be misleading as to the potency of a compound , and may call for validation in new animal model systems [59] . Furthermore , this study highlights the importance of using representative primary HIV-1 proteins for the purpose of vaccinology and drug discovery . Passage in culture , in the absence of pressure to maintain certain accessory gene functions , can lead to a lack of potency in several of these proteins , including Vpu . Thus , there are clear pitfalls that come with using potentially unrepresentative , albeit historically well characterised , proteins derived from laboratory-adapted viral strains such as NL4 . 3 .
Plasma samples were obtained from 14 treatment-naïve individuals ( i . e individuals that had not received treatment either during or prior to sampling ) enrolled at the Chicago Clinical Research Site for the Multicentre AIDS Cohort Study ( MACS ) . HIV-1 disease progression was defined by time to AIDS and included 5 long-term non-progressors ( defined as >10 years from seroconversion to onset of AIDS ) , 4 normal progressors ( 5–9 years to onset of AIDS ) , and 5 rapid progressors ( <5 years to onset of AIDS ) [32]–[34] . Where possible , a 1–2 year and 3–4 year post-infection time point was obtained per individual . Progression status was assigned retrospectively , as standard in the MACS . For the RPs in this study , early disease progression precluded the availability of a 3–4 year time point . For NP 3 and LTNP 4 , additional seroconversion time point ( 0 years ) was also analysed , and for LTNP 1 a 10 . 4-year time point . The volume of plasma equivalent to 10 , 000 copies of viral RNA ( based on standard clinical viral load measurements ) was first centrifuged to remove cellular debris ( 5 , 400×g at 4°C for 10 mins ) , then the virions were pelleted ( 25 , 000×g at 4°C for 1 hour ) . Viruses derived from blood samples collected in heparin were then heparinase treated to avoid inhibition of downstream enzymatic processes . RNA isolated from the virion pellets was then transferred to standard reverse transcription reactions ( Invitrogen Superscript III ) , using Vpu-specific outer reverse primer ( see SGA section below ) , according to manufacturer's instructions . Single genome amplification techniques were based on methods described in Palmer et al . [60] Nested PCR primers were designed to conserved regions in the tat/rev first exon and the env gene according to sequences derived form the Los Alamos National Laboratory HIV Sequence Database ( forward EK5846-5869 5′-CCT AGA CTA GAG CCC TGG AAG CAT-3′; reverse EK6473-6453 5′-TTC TTG TGG GTT GGG GTC TGT-3′ ) , with the inner primers containing standard sequencing primer sequences T7 and M13 ( forward EK5972-5990 5′-TAA TAC GAC TCA CTA TAG GCA GGA AGA AGC GGA GAC A-3′; reverse EK6848-6330 5′-CAG GAA ACA GCT ATG ACC CCA TAA TAG ACT GTG AC-3′ ) , numbered according to the HXB2 molecular clone . Viral cDNA was serially three-fold diluted from 1∶5 to 1∶405 and used as a template for multiple PCRs . We first performed 12–24 reactions at the highest dilution , and the number of positive reactions was used to calculate the cDNA dilution at which approximately 30% would be positive as predicted by the Poisson distribution . We then performed 92 PCRs at this modified dilution , and reactions yielding a product were directly sequenced with T7 and M13 primers ( MWG Eurofins , Germany ) . Chromatograms were carefully examined for the presence of double or multiple peaks . For each patient , complete vpu gene sequences were manually aligned with the software Se-Al version 2 . 0a11 [61] . Maximum likelihood phylogenies were reconstructed under the General Time reversible ( GTR ) model of nucleotide substitution , with gamma-distributed rate heterogeneity , using RaxMLGUI version 1 . 2 [62] . Robustness of the tree topologies was assessed by non-parametric bootstrap testing , with 1000 replicates , also performed with RaxmlGUI . Trees were edited using the software FigTree version 1 . 3 . 1 [63] . A maximum likelihood phylogeny containing sequences from all patients was also reconstructed following the same procedure . Intra-host pairwise genetic distances were calculated using the phylogenetic package HyPhy version 2 . 1 . 2 [64] . For each alignment , nucleotide and amino acid substitution matrices were estimated under the GTR and Whelan & Goldman models respectively . Codon-specific selection analyses were conducted via the HyPhy webserver DataMonkey [65] . Three different methods were used to identify vpu sites evolving under constant adaptive pressure: Single Likelihood Ancestor Counting ( SLAC ) , Fixed Effects Likelihood ( FEL ) and Fast Unbiased Bayesian Approximation ( FUBAR ) . For each patient , estimations were conducted under the best fitting model of nucleotide substitution , as selected by the model selection procedure implemented in DataMonkey . Sites showing evidence for positive selection by more than one method at the p<0 . 05 ( SLAC and FEL ) or posterior probability >0 . 95 ( FUBAR ) significance level were included in the study . In addition , the Mixed Effects Model of Evolution ( MEME ) method was used to identify sites subjected to episodic selective pressures ( posterior probability >0 . 95 ) . Sequences were screened for recombination prior to analyses , using the Single Breakpoint Recombination ( SBR ) and Genetic Algorithms for Recombination Detection ( GARD ) methods implemented in DataMonkey . No recombination breakpoint was found at the p<0 . 05 significance level . Codons found to be under positive selection that were located in the region of vpu that overlaps with the env open reading frame ( codons 55–81 ) , were assessed on an individual basis as follows: a non-synonymous change in vpu that was synonymous in env was scored as positive; a non-synonymous change in both genes was impossible to reliably determine which gene the selection was acting on , therefore these cases were excluded from the results . To identify codon-specific selective pressure on the vpu gene at the population level , the above-mentioned procedure was repeated on an alignment containing the unique nucleotide sequences from all patients ( n = 443 ) . For population-level selective pressure , data are presented pertaining only to the region of the vpu gene that does not overlap with the env open reading frame ( codons 1–55 ) . Vpu repertoires from each time point were stripped of duplicates , and all unique amino acid sequences from each sample were re-amplified using the inner forward and revers primers described above modified with EcoR1 and Not1 restriction sites respectively . Products from these reactions were then cloned into an Rev-dependent HIV-1-based expression vector , pCRVI [20] , to obviate the need to codon optimise the Vpu sequence [66] . The resultant plasmids were then re-sequenced , to ensure that no mutations were introduced into the vpu genes during the cloning process . HEK293T cells were seeded at 1 . 5×105 per well of a 24-well plate the day before transfection . Cells were co-transfected with 500 ng NL4 . 3delVpu provirus plasmid , or NL4 . 3 wild-type plasmid as a control , plus 50 ng pCR3 . 1-human tetherin plasmid , or pCR3 . 1 empty vector as a control , plus a standard input of 25 ng of pCRVI-Vpu , or pCRVI empty vector as a control . In the case of titration experiments , 5 , 10 , 25 , 50 and 100 ng of pCRVI-Vpu were used in each assay , with the total plasmid quantity kept constant by the addition of pCRVI empty vector plasmid to a total quantity of 100 ng . Cell culture medium was removed 14 hours after transfection , and replaced with 600 µl per well . 48 hours after transfection , viral supernatants and cell lysates were harvested , and infectious virus released determined by standard HeLa-TZMbl assay and virus particle release determined by Western blot . Each Vpu was tested in a minimum of three independent experiments , and results were compared between experiments by setting the level of infectious virus released in the presence of NL4 . 3delVpu virus plus pCRVI-NL4 . 3 Vpu as 100% , and expressing the activity of the patient-derived Vpus as a percentage thereof . NL4 . 3 Vpu constructs with defects specifically in tetherin antagonism ( A14L ) , or tetherin antagonism plus CD4 downregulation ( S52 , 56A and A14L/W22A ) were included in all assays as negative controls . HeLa-TZMbl cells were seeded at 8×104 cells per well of a 24-well plate the day before transfection . Cells were co-transfected with 150 ng pCR3 . 1-GFP or empty vector control , and 100 ng pCRVI-Vpu or empty vector control . 24 hours after infection , cells were harvested and stained for cell surface molecule expression using a mouse anti-human CD4 monoclonal antibody directly conjugated to allophycocyanin ( APC; clone RPA-T4; BD Biosciences ) , or a mouse anti-human tetherin monoclonal antibody ( clone 3H4 , Novus Biologicals ) followed by an IgG2a specific anti-mouse-Alexa Fluor 633 secondary antibody ( Life Technologies ) . Cells were then analysed for CD4 or tetherin and GFP levels using a FACSCalibur flow cytometer ( Becton Dickinson ) and FlowJo software ( TreeStar Inc , Oregon , USA ) . Cells expressing high levels of GFP were gated and CD4/tetherin levels were determined as the median fluorescent intensities . Absolute downregulation levels were calculated as the percentage reduction of CD4/tetherin cell surface expression ( median fluorescent intensity ) in the presence of Vpu compared to in the absence of Vpu ( empty vector transfection ) . For the purposes of comparison with the tetherin counteraction assay , the absolute level of CD4/tetherin downregulation obtained in the presence of NL4 . 3 Vpu was normalised to 100% , and therefore the CD4/tetherin downregulation by all other Vpus expressed as a percentage thereof . ( Note that the absolute CD4 downregulation in the presence of NL4 . 3 Vpu was 80+/−6% ) . HEK293 cells were seeded at 1 . 2×105 per well of a 24-well plate the day before transfection . Cells were co-transfected with 10 ng 3×κB-pConA-FLuc 50 and 5 ng pCMV-RLuc reporter constructs , plus pCR3 . 1-human tetherin plasmid , or 3 . 1-MAVS/IPS1/Cardif , or GFP plasmid as a control , and 50 ng of pCRVI-Vpu or pCRVI empty vector as a control . 24 hours after transfection cells are harvested and luciferase activity measured with the Dual Luciferase Reporter Assay System ( Promega ) . Luciferase signals were normalised , and fold NF-κB activation calculated in the absence of Vpu expression . In the case of titration experiments , 5 , 10 , 20 , 50 and 100 ng of pCRVI-Vpu were used in each assay , with the total plasmid quantity kept constant by the addition of pCRVI empty vector plasmid to a total quantity of 100 ng . All Vpus classified as defective or suboptimal for both CD4 downregulation and tetherin counteraction ( i . e . 0–75% that of NL4 . 3 Vpu activity ) were tested for expression in 293T cells by Western blot analysis using a polyclonal rabbit anti-Vpu antibody [67] kindly provided by Klaus Strebel through the NIH AIDS Reagent Program . Since this antibody is specific for the C-terminal region of NL4 . 3 Vpu , and the patient-derived Vpus differ considerably in amino acid sequence from NL4 . 3 Vpu , expression of each defective/suboptimal Vpu was compared to that of its nearest functional relative from the same infected individual . Defective/suboptimal Vpus that showed low or no expression were re-transformed , re-purified and then re-sequenced to ensure the quality of the plasmid preparation , and in all cases the expression levels before and after this process were comparable . Defects were therefore deemed to be due to natural expression defects or instability of the expressed protein . 293T cells were co-transfected with 600 ng of pCR3 . 1-β-TrCP2 myc or GFP and 600 ng of pCRVI-Vpu or EV . 48 hours after transfection , cross-linking immunoprecipitations were performed [68] . Briefly , cross-linking was preformed on harvested cells using 0 . 05% HCHO , then lysed in 150 mM NaCl , 10 mM Hepes ( pH ) , 6 mM MgCl2 , 2 mM DTT , 10% glycerol , 0 . 5% NP40 , 200 µM sodium orthvanadate and protease inhibitor cocktail . Cleared lysates were immunoprecipitated with mouse anti-myc monoclonal antibody ( clone 9E10 , Covance ) and protein G agarose beads ( Invitrogen ) . Cross-linking was reversed with 10 mM EDTA , 5 mM DTT and 1% SDS , and lysates and immunoprecipates were analysed by Western blot using mouse anti-myc and rabbit anti-Vpu antibodies . Unpaired two-tailed T tests were used to determine significant differences between samples for the CD4 downregulation ( Figure 2A ) , tetherin counteraction ( Figure 2B ) and suppression of tetherin activation of NF-κB ( Figure 6 ) . A two-tailed Fisher's exact test was used to determine whether Vpus containing a threonine or valine at position 15 instead of alanine decreased over time in certain individuals . Levels of significance were determined as follows: *** p<0 . 001 , ** p<0 . 01 , *p<0 . 05 , ns p>0 . 05 . Anonymized , pre-collected plasma samples and associated clinical data used in this study were obtained from the Chicago MACS Center with the permission of the Multicenter AIDS Cohort Study/Women's Interagency HIV Study . Permission to use anonymized human plasma samples was also granted by the King's College London Infectious Disease BioBank Local Research Ethics Committee ( SN-1/6/7/9 ) . Data in this manuscript were collected by the Multicenter AIDS Cohort Study ( MACS ) with centers ( Principal Investigators ) at The Johns Hopkins Bloomberg School of Public Health ( Joseph B . Margolick ) , Northwestern University , and Cook County Bureau of Health Services ( Steven Wolinsky ) , University of California , Los Angeles ( Roger Detels ) , and University of Pittsburgh ( Charles Rinaldo ) . The Data Center is located at the Johns Hopkins Bloomberg School of Public Health ( Lisa P . Jacobson ) . The MACS is funded by the National Institute of Allergy and Infectious Diseases , with additional supplemental funding from the National Cancer Institute . UO1-AI-35042 , UM1-AI-35043 , UO1-AI-35039 , UO1-AI-35040 , UO1-AI-35041 . Website located at http://www . statepi . jhsph . edu/macs/macs . html .
|
The accessory protein Vpu , encoded by HIV-1 , performs at least two major roles in the virus life cycle , namely the degradation of newly synthesized CD4 molecules and the counteraction of a host antiviral protein , tetherin . These activities promote the release of infectious viruses from host cells , and recent evidence suggests that Vpu function has been crucial for the cross-species transmission of HIV-1 from chimpanzees , and its subsequent pandemic spread in humans . Here we studied the functional variation in Vpu in infected individuals . We found that the Vpu amino acid sequence can be highly variable within an individual , and that this variation is likely to result from host immune responses targeting antigens derived from Vpu . However , despite this variation , Vpu's major functions are preserved , with only a minority of circulating alleles showing defects throughout the course of infection . These data suggest that defective Vpu proteins are selected against within the infected individual , implying that Vpu functions are critical for HIV-1 replication throughout natural infection , not simply at transmission . Therefore Vpu may represent a novel target for antiviral therapy to augment current treatment strategies for HIV/AIDS .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"hiv",
"virology",
"retrovirology",
"and",
"hiv",
"immunopathogenesis",
"biology",
"microbiology",
"viral",
"diseases"
] |
2014
|
Preservation of Tetherin and CD4 Counter-Activities in Circulating Vpu Alleles despite Extensive Sequence Variation within HIV-1 Infected Individuals
|
The neuronal code arising from the coordinated activity of grid cells in the rodent entorhinal cortex can uniquely represent space across a large range of distances , but the precise conditions for optimal coding capacity are known only for environments with finite size . Here we consider a coding scheme that is suitable for unbounded environments , and present a novel , number theoretic approach to derive the grid parameters that maximise the coding range in the presence of noise . We derive an analytic upper bound on the coding range and provide examples for grid scales that achieve this bound and hence are optimal for encoding in unbounded environments . We show that in the absence of neuronal noise , the capacity of the system is extremely sensitive to the choice of the grid periods . However , when the accuracy of the representation is limited by neuronal noise , the capacity quickly becomes more robust against the choice of grid scales as the number of modules increases . Importantly , we found that the capacity of the system is near optimal even for random scale choices already for a realistic number of grid modules . Our study demonstrates that robust and efficient coding can be achieved without parameter tuning in the case of grid cell representation and provides a solid theoretical explanation for the large diversity of the grid scales observed in experimental studies . Moreover , we suggest that having multiple grid modules in the entorhinal cortex is not only required for the exponentially large coding capacity , but is also a prerequisite for the robustness of the system .
Optimising neuronal systems for efficient processing and representation of information is a key principle for both understanding and designing neuronal circuits [1] , but deciding whether a particular neuronal phenomenon reflects an optimisation process is often difficult . Grid cells in the medial entorhinal cortex have been suggested to efficiently represent spatial location of the animal by their spatially periodic firing fields near optimally [2 , 3 , 4 , 5] . However , it remained controversial whether the efficiency of the grid cell code is the result of the precise tuning of the grid parameters [6 , 7 , 8] or the performance of the system is relatively insensitive to the actual parameter settings [4 , 5 , 9] . Grid cells are spatially tuned neurons with multiple firing fields organised along the vertices of a triangular grid ( Fig 1a; [10 , 11] ) . Grid cells of any particular animal are organised into functional modules [12 , 13] cells within a module share the same grid scale and orientation , but differ in the location of their firing fields , i . e . , their preferred firing phase within the grid period ( Fig 1a ) . Modules form the functional units of the grid representation: The joint activity of all ( possibly hundreds of ) cells within each module is captured by the ( two dimensional ) phase of the given module ( Fig 1b; [14 , 15] ) and the relationship between different cells from the same module remains stable across different environments [16] , during sleep [17 , 18] or after environmental distortions [13] . A given spatial location is represented by the phases of the different modules ( ‘phase vector’ ) . The representations are unique up to a critical distance above which the coding becomes ambiguous: the phase vectors , and hence the firing rates of all grid cells , become ( nearly ) identical at two separate physical locations ( Fig 1c ) . Depending on the magnitude of the critical distance compared to the largest grid scale , two complementary coding schemes have been proposed for grid cells ( Fig 1c ) : In nested coding [4 , 6 , 8] smaller grid modules iteratively refine the position coding of larger modules and the modules span a wide range of scales . The capacity of nested codes , defined as the ratio of the coding range and the resolution , is exponential in the number of modules . Maximal capacity can be achieved by setting the coding range equal to the maximal grid period and then optimising the resolution by a geometric progression of the grid scales [6] . When the total capacity is utilised to encode locations within the maximal grid scale , catastrophic interference will cause ambiguity in the grid code beyond this distance ( Fig 1c ) . When the coding is not optimised for a fixed range , the unique combination of the activity of grid modules can encode a potentially unbounded range that can be substantially larger than the scale of the largest module using a modulo arithmetic ( MA ) code [2 , 3 , 14] ( Fig 1c ) . In this case the grid periods can be similar in magnitude ( e . g . , co-prime integers or a geometric progression with a relatively small ratio ) . However , it is not known under what conditions the MA coding system can achieve exponential capacity [3 , 14] , and how robust is the capacity against the choice of the grid periods or neuronal noise . Here we develop a novel approach to study the capacity of the grid coding system that is based on Diophantine approximations , i . e . , approximation of real numbers by rational numbers . First , we apply the technique to study coding with two grid modules . We show that the capacity of the system is extremely sensitive to the number theoretic properties of the scale ratio between the modules . Next , we generalise our approach to the case of multiple modules , and show both analytically and numerically that the exponential capacity of the grid cell coding system can be achieved using the MA coding scheme . Finally , we demonstrate that when the coding range is constrained by neuronal noise , the capacity of the system is extremely robust against the choices of the scaling of the modules .
We can formalise the problem of interference between two modules as having a pair of integers k and ℓ with ℓ ≈ kα , meaning that module 2 ( with scale α ) is close to being in phase 0 at distance ℓ , which would cause ambiguity between the coding of the spatial point ℓ and the origin . This is formally identical to the number theoretic question of the approximability of the scale α ≈ ℓ/k with rationals having numerator ℓ , also known as Diophantine approximations ( Fig 2b and 2c ) . Hurwitz’s theorem [19 , 20] states that for all irrational numbers α > 1 there are infinitely many relative primes k , ℓ such that the error of the approximation , defined as ϵ ( ℓ ) = | k - ℓ / α | , ( 2 ) is smaller than the upper bound: ϵ ( ℓ ) < 1 5 1 ℓ . ( 3 ) Note that the approximation error ϵ ( ℓ ) ( Eq 2 ) is the same as the phase difference between the modules , defined in Eq 1 , since ψ2 ( ℓ ) = [ℓ/α] mod 1 = |k − ℓ/α| for an appropriately chosen integer k ( Fig 2b and 2c ) . We call ϵ ( ℓ ) ‘approximation error’ only when we are talking about approximating irrationals with integer ratios while in the context of grid cells we will call ϵ ( ℓ ) the ‘phase difference’ . Applied to the grid cells , Hurwitz’s theorem provides an upper bound on how the phase difference between the modules shrinks with the distance . Specifically , the theorem states that there are infinitely many integer distances ℓ , where the phase difference is smaller than ϵ ( ℓ ) ∝ 1/ℓ ( Fig 4a–4c , dashed lines ) , implying that on the long run interference can not be avoided no matter how carefully we choose α . This is a fundamental upper bound on the efficiency of coding with grid cells . The critical distance where the phase difference ϵ ( ℓ ) leads to interference , and the representation of the position becomes ambiguous , depends on the noisiness of the two modules , δ and αδ . Interference occurs if there is a spatial point x for which both |x − kα| < αδ and |x − ℓ| < δ for integers k , ℓ , or equivalently , if |kα − ℓ| < ( 1 + α ) δ . Hence , by the definition of ϵ ( ℓ ) ( Eq 2 ) the coding is ambiguous near ℓ if and only if ϵ ( ℓ ) < 1 α ( 1 + α ) δ . ( 4 ) Therefore , no matter how we chose α , we can expect ambiguity at distances ℓ from the origin if the noise in the system is larger than the upper bound on efficiency provided by Eq 3 , i . e . ℓ > α 5 ( 1 + α ) 1 δ , ( 5 ) that is , at distance of order 1/δ . Consequently , it is impossible to code position with two modules better than this bound . The question arises then whether the above theoretical bound is achievable , at least for some appropriately chosen α . The answer is yes , namely the upper bound in Eq 3 is sharp for the golden ratio α = σ ≔ 5 + 1 2 ≈ 1 . 618 . Practically , this also introduces a limit on ϵ ( ℓ ) , saying that the phase difference between the modules remains always larger than a specific lower bound: ϵ ( ℓ ) > ( 1 5 - ε ) 1 ℓ ( 6 ) except for a couple of small distances , even for arbitrary small ε > 0 [19] ( Fig 4b ) . It may sound strange that there are finitely many exceptions , but in our simulations we found only a few instances with ℓ being small ( Fig 4e ) . Therefore , if the ratio of the two grid modules equals the golden ratio then the phase difference between the two modules is guaranteed to be larger than the lower bound defined by Eq 6 . Since ε can be arbitrarily small , the lower bound for the golden ratio approaches the theoretical upper bound ( Eq 3 ) and σ is an optimal choice for the scale ratio to avoid interference in case of two modules . To give a geometric picture , the golden ratio guarantees approximately uniform coverage of the phase space for both short and arbitrarily long distances ( Fig 2c , right ) . However , it turns out that there are many good choices [20]: for any algebraic integer α of order 2 ( i . e . irrational which is a root of a polynomial of degree 2 with integer coefficients , see Methods ) there exists a maximal positive constant cα > 0 such that ϵ ( ℓ ) ≥ c α 1 + α α 1 ℓ ( 7 ) holds except for a couple of small distances ( Fig 4c–4f , [21] ) . Hence , from Eqs 4 and 7 we see that the representation is unambiguous whenever cα ( 1 + α ) / ( α ℓ ) > δ ( 1 + α ) /α , that is up to ℓ ≤ L max ≔ c α δ ( 8 ) for all δ which is small enough . This last condition on the magnitude of the noise is only needed to exclude the possible exceptionally small ℓ distances in Eq 6 , which in practice is not a crucial condition ( Fig 4d and 4e ) . The constant cα is the single parameter that determines the critical distance up to which encoding is unique ( coding range ) as well as the information rate of the system ( Methods ) . Therefore , we use cα to compare the efficiency of different choices of α ( Fig 4d–4f ) . We have already noted that for the golden ratio the lower and the upper bounds ( Eqs 3 and 7 ) coincide ( Fig 4b ) , but the critical distance may be larger for some α even if the corresponding lower bound on the phase difference is weaker , since the upper bound also depends on α ( Eq 5 ) . We estimated the value of cα for various scale ratios at different noise levels ( Methods ) . Unlike for algebraic numbers , c ^ α of real numbers depends on the distance range used for the estimation , which we controlled by setting different intervals for δ in the simulations . Our simulations confirmed that σ is the best scale ratio choice in case of two modules with c ^ α = 0 . 28 ≈ σ 5 ( 1 + σ ) , but also showed that , on both short and long run , c ^ α is extremely sensitive to the choice of α ( Fig 5 ) : in case of a small error in the tuning of α , the efficiency can drop substantially and c ^ α becomes practically 0 , implying that in the immediate neighbourhood of the optimal α , there are close to pessimal grid cell configurations . This is because the lower bound on the phase difference ( Eq 7 ) requires α to be an algebraic number , and in an arbitrary small neighbourhood of any algebraic number there are ( infinitely ) many non-algebraic numbers , i . e . , transcendental numbers ( α = e , Fig 4a ) or rational numbers ( α = 3/2 , cα → 0 , Fig 2a ) . As non-algebraic irrational numbers can be much better approximated with rationals than algebraic numbers , non-algebraic grid scale ratios will lead to much stronger interference between the two modules , but only at distances moderately large compared to the scale of the modules ( Fig 5 ) . The extremely rough landscape of cα renders optimisation for α an especially difficult problem: it is very unlikely that a biological system would be able to find the global optimum for the scale ratio of two grid modules and a relatively small mistuning from a local optimum could significantly deteriorate the efficiency of the system . Therefore , at least in the case of two modules , it seems to be impossible to achieve asymptotically optimal scale ratio for the grid cells . To derive the general solution for M grid modules , we focus on a set of 1-dimensional grids with scales α0 = 1 < α1 < ⋯ < αM−1 . Spatial representation is unambiguous up to a distance L from the origin if there is at least one module for which the phase is significantly different from 0 ( Fig 6 ) . Interestingly , avoiding interference between adjacent modules ( giving α = σ ) is not a good solution , since it leads to interference between the distant modules wherever the adjacent modules are in close apposition ( Methods ) . The logic of the general solution for multiple modules is the same as in the case of two modules . Here we only state the main results and the technical details of the analysis can be found in the Methods . First , we show that a similar upper bound exists for the maximal phase difference between the modules . Compared to the two-module case , the bound is weaker when M > 2 as the phase difference scales only with 1/ℓ1/ ( M−1 ) ≫ 1/ℓ meaning that it ensures simultaneous interference between all modules only at much larger distances . Second , we found that the upper bound can be satisfied , up to a constant multiplier , c A , for algebraic scale ratios ( Methods ) . Specifically , if the scales of M modules form a geometric series with common ratio α being an algebraic number of degree M , the upper bound is tight , meaning that the phase difference does not shrink faster than 1/ℓ1/ ( M−1 ) . Intuitively , this scaling indicates that there is always at least one pair of modules for which the phase difference at the integer distance ℓ from the origin is larger than the lower bound . The critical distance Lmax up to which coding is unambiguous can be expressed as ( cf . Eq 8 ) : L max ≔ ( c A δ ) M − 1 , ( 9 ) for all δ which is small enough , where c A and its estimate c ^ A are defined analogously as in the two modules case ( see Methods for the definition ) . Intuitively , Eq 9 expresses an exponential scaling of the maximal distance uniquely represented by a population of grid cells with the number of grid modules , M . The coding range of a particular set of the grid scales , A = ( α 1 , … , α M - 1 ) , depends on both the noise in the system and on the basis of the exponential c A . Interestingly , the above described geometric sequence of algebraic numbers are the only known explicit examples of badly approximable vectors ( to the best of our knowledge ) . However , it is known that there are much more such vectors which do not form geometric sequences [22] , therefore the scale ratio of a well-tuned MA grid cell system does not have to be constant . The expression about the exponential scaling ( Eq 9 ) is similar to the capacity estimates of Fiete et al . [3] ( see their Eq 6 ) obtained using a combinatorial upper bound and numerical simulations . Importantly , our analytic derivation also provides insight about why certain grid systems are more efficient than others and give examples for efficient grid scales . Moreover , when c A = 0 . 5 our formula for the capacity of the grid code becomes identical to the theoretically maximum capacity found in the case of nested coding [4 , 6] . In the next sections we first numerically estimate the value of c A for various choices of the grid scales A and then we show that with sufficiently large number of modules c A is guaranteed to approach its theoretical maximum c A = 0 . 5 for randomly chosen grid periods . We developed an efficient method to numerically estimate the value of c A for various parameter settings that is based on the simultaneous Diophantine approximations of a set of irrational numbers ( Methods ) . Using realistic noise levels we found that , in contrast to the case of two modules , the sensitivity of the coding efficiency to the choice of α gradually vanishes when the number of grid modules is increased ( Fig 7a and 7b ) . In particular , with M = 10 modules c ^ A ∈ [ 0 . 2 , 0 . 4 ] for almost all choices of the grid scales ( Fig 7b ) , both when the scales follow a geometric series with a common scale ratio α ( Fig 7b ) and when all the M scales are chosen from the bounded interval ( 1 , 2 ) . We also found that c ^ A vanishes only for pathological examples such as rational numbers or powers of the second order algebraic number α = σ − 1/2 ≈ 1 . 118 ( Fig 7a and 7b , red ) . The only random scale choice that significantly degrades the performance is when α ≈ 1 ( Fig 7a and 7b ) in which case all grid modules have nearly identical spatial scale . To quantify the sensitivity of the grid system against the choice of the scale parameters we calculated the mean and the coefficient of variation of c ^ A with random choices of α ( Fig 7c and 7d ) . We found , that the average c ^ A increased monotonically with the number of grid modules indicating that the system’s performance becomes closer to the ideal c A = 0 . 5 value as the number of modules increased ( Fig 7c ) . Moreover , the variability of c ^ A consistently decreased with the number of modules reflecting the improved robustness of the system to the choice of grid periods ( Fig 7d ) . Therefore not only the maximal coding distance increases exponentially with the number of modules , but the basis of the exponential , c A , also increases . To further investigate the mechanisms responsible for the robustness of the system , we numerically evaluated the minimal phase difference between the modules , ϵ ( ℓ ) , in the function of the distance ( Fig 7e and 7f ) . In line with the predictions of the theory ( Eq 25 ) , we found that the phase difference decreased with ℓ−1/ ( M−1 ) , i . e . , with a small negative power of the distance for α being an order M algebraic number ( Fig 7e , black ) . For suboptimal α-s , the scaling of the phase difference was nearly optimal up to a critical point beyond which the scaling followed the algebraic rank of α ( i . e . , second order α scales with 1/ ℓ , Fig 7e , yellow ) . Importantly , this critical point , where the transition occurs between ideal and number theoretical scaling is located at increasingly larger distances when the number of modules is increased ( Fig 7f ) . Therefore , the asymptotic , number theoretical properties of the grid periods have a gradually lower impact on the performance of the system in the distance range limited by the intrinsic variability on neuronal spiking ( Fig 7e and 7f , background shading ) . These observations suggest that even random scale choices might achieve optimal performance as the number of modules grow . In the next section we make this statement mathematically precise and demonstrate that indeed , c ^ A approaches its theoretical maximum , 0 . 5 , when the number of modules grow and the scales are chosen uniformly at random from a bounded interval . Our number theoretic argument ( Eq 9 ) alone does not imply exponential capacity , since it does not exclude the possibility that the base of the exponential , c A , converges to 0 as M increases ( although we observed the opposite trend , see Fig 7c ) . In this section we investigate the asymptotic properties of the grid code when the number of modules increases and the relative uncertainty δ of the modules remains fixed . Here we only state these results informally , and leave the precise statements and the slightly technical mathematical proof to the Methods . The main idea behind the proof is that the phase of a given module at particular distance x from the origin depends only on the scale of that module , α . If the scale is randomly chosen from a bounded interval [1 , αmax] , then the phase is also a random variable with probability distribution approaching the uniform distribution as the distance increases . Then , the probability of simultaneous interference between M modules , that is , the probability of all modules being near phase 0 at some distance x , is proportional to the volume of an M-dimensional hypercube , which is V = ( 2δ ) M , where the side of the cube is 2δ . The ratio of the volume of the hypercube and the unit cube ( the number of distinguishable phases ) diminishes exponentially with M , and the total distance ( expressed in units of α0 = 1 ) covered without ambiguity is 2 δ V ∝ ( 1 2 δ ) M - 1 . Specifically , our statement is , roughly speaking , that if 0 < δ < 1/2 is fixed , M is large enough , and the module scales are drawn uniformly at random from a not too narrow bounded interval , e . g . from ( 1 , 2 ) , then the representation is unambiguous up to the exponential distance l < ( c A δ ) M − 1 ( 10 ) with probability approaching 1 , and c A approaching 1/2 . Although the above statement applies only for M → ∞ , and it does not provide examples for efficient scale choices for finite M , we emphasise that this result is stronger than our previous derivation ( Eq 9 ) in four aspects: First , our previous derivation ( Eq 9 ) allowed c A to tend to 0 as M increased . Now we showed that this does not happen for random scale choices , rather the value of the constant tends to its theoretical maximum , c A = 0 . 5 [6] for large M with high probability , confirming our previous numerical results ( Fig 7c ) . Second , one can achieve this nearly optimal performance without increasing the scales exponentially , with the scales chosen from a bounded interval . Third , this almost optimal efficiency is not only reached for some appropriately chosen scales , but for almost all choices . Fourth , near-optimal performance is guaranteed for 2 or higher dimensional grid codes even if the modules are randomly rotated relative to each other or in the absence of long-range coherence within the modules . Thus , our results demonstrate that no meticulous tuning of the grid scales is required for close to optimal grid system performance .
Previous works used specific assumptions to derive exponential coding range for the grid cell coding system: they assumed either a nested coding scheme [5 , 6] or presumed that the phase space is covered evenly and that the readout noise in a given module decreases when the number of modules increases [3 , 14] . Here we generalised these findings and demonstrated that nested and MA codes have asymptotically equal capacity . When we studied the capacity of MA codes we realised that achieving uniform coverage of the phase space is not trivial in the case of two modules , but can only be attained with appropriately chosen scales . Specifically , we recognised that approximately uniform coverage of the phase space by the phase curve at arbitrary distances is guaranteed if the scale ratio between the two modules is an algebraic number of order 2 . Using our formalism allowed us to generalise this intuition for arbitrary number of grid modules and to demonstrate that even a random choice of grid scales guarantees uniform coverage of the phase space when the number of modules is high . We also relaxed the assumption of an earlier study [14] that the total amount of the noise remains constant in the grid system even when the number of modules is increased , i . e . , the readout error of each module decreases with M . Here we derived these results using the more general assumption that the coding precision of each module is independent of M and proportional to the scale of the module . We confirmed our analytical results by extensive numerical simulations regarding the simultaneous interference between grid systems with various choices of the scale parameters . In line with previous results [3 , 9] , our simulations supported that the grid system is robust to the choice of the scale parameter and that the coding range is exponential in the number of modules . Although the efficiency of the coding investigated in this paper is slightly worse than that of the optimal nested coding [5 , 6] , MA codes also have several advantages . First it uses orders of magnitude smaller scale lengths than the maximal distance up to which the coding works properly . The largest grid scales measured experimentally are ∼3 m [23 , 24] and extrapolations based on the dorso-ventral location of the recording electrodes within the entorhinal cortex extend to ∼10 m [13] , a period still substantially smaller than the typical distances travelled daily by rodents ( several hundreds of meters [25] ) or bats ( several kilometres , [26]; see also [27] ) . Second , while the consequence of a module failure simple decreases the capacity of the system in the case of MA coding , it can have more dramatic effect in nested codes: Although malfunction of the largest or smallest module reduces either the capacity or the resolution of nested codes , respectively , the lack of intermediate modules functionally breaks the interaction between the remaining modules decreasing both the resolution and the capacity of the system in a disproportionate manner . Third , once the scales are optimised for a given noise level , the coding range of nested grid codes does not depend on δ . Therefore , contrary to MA codes , it is not possible to increase the capacity by inserting more neurons into the same modules or by observing more grid cells from the same set of modules . Conversely , the functioning of the nested codes critically depends on accurate decoding of each module: If the readout neuron does not have access to enough presynaptic neurons from a given module , then the corresponding posterior becomes too wide leading to interference between the modules . This has similar consequences as the absence of the given module in nested codes . In contrast , in MA codes the coding properties remain similar for postsynaptic neurons receiving different number of synapses from different modules , although the coding range is the function of the precision available for the observer ( Eq 9 ) . When encoding dynamic trajectories instead of static locations , the number of neurons required to participate in a given module decreases quadratically with the scale of the module , i . e . , n i ∼ 1 / α i 2 [8] . For example , if representing the position in the 2D space with some fixed accuracy with αi = 0 . 2 m requires ∼ 4000 neurons then αj = 2 m needs only ∼ 40 neurons . This scaling implies that the coding range of the nested grid system can be easily and parsimoniously extended by adding a new module with larger scale but containing only relatively few neurons . Although the relationship between the number of neurons in a module and its scale holds also for MA codes , the total number of neurons required to achieve similar coding range can be substantially smaller in nested codes . Another consequence of dynamical coding is that the time constant of the readout has to be matched to the scale of the grid modules [8] . As the grid scale varies over a large range in the case of nested codes , the postsynaptic neuron has to integrate inputs from different grid cells with time constants ranging from 1 ms to 1 second [8] . In MA codes , the modules have similar scales and their outputs can be integrated with similar time constants . Finally we note that nested coding and MA coding are not mutually exclusive: although they imply fundamentally different way of decoding the same positional information [7 , 14] , but both can be present in the same system . The MA code has a larger coding range if c A > α δ so it is favoured by small α ( small differences between scales ) and small δ ( high accuracy ) . Even in this case locations within the largest grid scale can be decoded as in nested coding , while MA decoder is required beyond this distance . In the Methods we show that the coding capacity of two or higher dimensional grid cells depend on the same number theoretic properties , and therefore the results obtained in dimension one extend to planar or cubic grid cells as well [28] , provided that the main axes of the different modules remain aligned with each other . If the two dimensional grid modules are rotated compared to each other , then the scale choices which perform well will be different from the scales that are optimal for axis aligned modules . Consider for example that α = 3 , which is a relatively good choice for M = 2 ( rightmost red circle in Fig 5a–5c ) , leads to cathastrophic interference at ℓ = α when the grids are rotated by 30° . Consequently , the incoherent reorientation of the grid modules during global remapping [13] renders the optimisation of the grid scales unfeasible . However , the main point of this paper is that we have shown analytically that almost all scale choices perform near optimally if the number of modules is high enough , which also applies for grid cells rotated uniformly at random relative to each other ( Methods ) . Moreover , the 2D grids does not need to show perfect triangular symmetry to achieve high capacity: environmental boundaries [29 , 24] or non-euclidean geometry [30 , 31] can distort the grid pattern , but as long as the distortion is coherent among modules , our theory applies unchanged . If the scales slightly vary on the long range , then our derivation based on the Diophantine approximations does not apply . However , our derivation stating exponential capacity for grid systems with many random scales ( Methods ) remains still valid . The highly organised , regular patterns formed by the firing fields of grid cells suggest that the characteristics of the grids must be closely related to the computational function of these neurons: optimally representing and processing information about the spatial location of the animal [32 , 33 , 14 , 4 , 11] . Besides the general optimality of triangular grid-like firing fields for representing unbounded 2D space [28] , recent theoretical work derived optimal scale ratio of successive grid modules in the case of nested coding [6 , 7 , 8] . These studies , using different assumptions , arrived at slightly different conclusions regarding the optimal value of α . Stemmler et al . , [7] fixed both the coding range to Lmax = 3 for a pair of grid modules with scales {1 , α} and found that α = 3/2 minimises the ambiguity errors within that range . Mathis et al . , [4] and Wei et al . , [6] also fixed the coding range and minimised the number of neurons required to achieve a given resolution and provided both estimates for the maximal capacity of the grid cell coding system and a specific architecture ( i . e . , optimised nested codes ) that achieves maximal efficiency . The optimal scale ratio for nested codes was found to depend both on the magnitude of the noise in the system and on the type of decoder [4 , 6] . Rather than fixing the coding range , we were interested in grid codes that work for potentially unbounded environments and found a similar asymptotic capacity for MA codes using random grid scales . Although predictions derived from nested coding roughly agree with the average scale ratio observed in the entorhinal cortex [12 , 13 , 29] , they do not explain the substantial amount of variability which characterises the data . In our derivations we assumed that the decoding error of a given grid module is larger than δ with some small probability . Inaccurate decoding of a single grid module can lead to disproportionally high error in the position representation if the subsequent time frames are decoded independently [14 , 9] . However , the chance of catastrophic ambiguity errors can be substantially reduced if a dynamical decoder combines prior information representing the predicted spatial position with the location encoded by the incoming grid cell spikes [34 , 14 , 8] . Our results based on the Diophantine approximations requires that the scale of the modules are set precisely , so that the phase of the different modules does not drift relative to each other ( i . e . , α i d ϕ i d x = α j d ϕ j d x ) . Although theoretical considerations suggest that drift can not be completely suppressed in a noisy neuronal system [35 , 36] , whether different grid modules respond coherently to distortions caused by environmental manipulations is not known [15 , 29 , 24] . The remarkable robustness of the grid system’s efficiency against the choice of the scale ratio suggests that grids with loosely set scale parameters could also obtain a similar performance . Indeed , our derivation using randomly selected grid scales does not require precisely set scale parameters yet it provides the asymptotically exponential capacity for the grid system . The optimization principle assumes that substantial improvement in the performance of the system can be achieved with precise tuning of its parameters . In the present study we demonstrated that this is indeed the case in the absence of noise . However , even in this case , optimization would be almost unfeasible for three reasons . First , the coding range is an extremely irregular , discontinuous function of the scale parameter , making optimisation essentially a trial and error game . Second , a scale parameter that is optimal for a given number of modules is guaranteed to be inefficient when the number of modules is increased precluding the possibility of pairwise or modular optimization . Finally , the optimal grid scales depend on the rotation of the modules relative to each other , which can change independently during changes in the environment [13] . However , taking the variability of neuronal firing into account changes the picture dramatically . We demonstrated that when the coding accuracy of grid modules is limited by neuronal noise , the capacity of the system becomes surprisingly robust to the choice of the scale parameters making its optimization unnecessary . Note , that even if the grid periods are not optimized across modules , generating the regular , periodic firing fields of grid cells demands accurate integration of velocity inputs [37 , 36] and repeated error correction [38 , 35] , both requiring the precise tuning of single neuron and network parameters within a given module . In conclusion , our study demonstrates that the capacity of the grid cell system is nearly optimal with randomly chosen grid scales , and , instead of accurate parameter tuning , the experimentally observed scales could reflect the combined effect of random fluctuations and a gradient in the cellular properties along the dorso-ventral axis of the entorhinal cortex [39 , 40] . Our finding , that grid cells have an exponentially large coding range even with randomly chosen grid scales of similar magnitudes makes several important predictions . First , MA coding predicts that the coding range is substantially larger than the largest grid period . Since grid cells are likely to be involved in path integration [32 , 41] this prediction could be tested by probing path integration abilities of rodents beyond distances of the largest grid period [42] . Second , in the case of MA coding , different modules have similar contributions to the coding range of the system . Therefore , the effect of targeted dMEC lesion ( inactivating a single module , as in [43] ) on the rat’s navigation behaviour would be largely independent of the actual location of the lesion ( i . e . , which module is inactivated ) . Third , since the performance of the system is independent of the precise choice of the grid scales , we expect a large variability in the scale ratio of successive grid modules both within and across animals . This prediction is consistent with the experimental data available [12 , 13 , 29] , although further statistical analysis would be required to specifically determine the distribution of scale ratios . Finally , we predict that the performance of the system is not particularly sensitive to incoherent changes in the scale parameter of a subset of modules during e . g . , global remapping induced by environmental changes [16] . It has been shown that under certain conditions simultaneously recorded grid cells respond coherently within a module and independently across modules to environmental distortions [13] . To test the prediction of our theory , the behavioural consequences of incoherent realignment across modules should be assessed and compared with the effects of environmental manipulations inducing coherent realignment [29] or coherent distortion in the shape of the grid pattern [24 , 29 , 30 , 31] .
Consider a system G2 of planar grid cells with a set of scales A . Suppose that the axis of all modules are aligned and use the coordinate system B = { [ 1 , 0 ] , [ cos ( 60 ° ) , sin ( 60 ° ) ] } , ( 11 ) which is naturally generated by the triangular lattice . To compare with consider the one dimensional grid cell system G1 which has the same number of modules with the same set of scales , and for which each module represents the position of the animal with the same relative precision . To achieve this , the two dimensional modules need squared as many cells , nevertheless they also able to distinguish between squared as many spatial positions within one period of the scale . If G2 represents a planar position ( x , y ) B ambiguously , i . e . , ψ ( x , y ) B ≈ ψ ( 0 , 0 ) B = 0 , then clearly planar positions ( x , 0 ) B and ( 0 , y ) B are also represented ambiguously . Therefore , the corresponding one dimensional positions x and y are represented ambiguously by G1 as well . Conversely , if z is represented ambiguously by G1 , then ( z , 0 ) B , ( 0 , z ) B will be ambiguous in G2 . Therefore , an ambiguity of position at a given distance from the origin in case of planar cells can be matched to an ambiguity at the same order of magnitude of distance in the one dimensional grid system , and vica versa . The above argument also shows that the same scale choices perform best for both one dimensional grid cells and two or higher dimensional ones when the axes are aligned with each other . We chose α0 = 1 and fix the resolution of the system to δ < α0 ( defined below ) and investigate its coding range . A formally identical system with a fixed coding range and optimised resolution can be achieved by appropriately rescaling the grid scales . We numerically estimated the precision of position coding by a single module by first simulating the motion of the animal as a one dimensional Gaussian random walk: P ( x t + 1 | x t ) = N ( x t , Δ t D ) ( 12 ) with Δt = 1 ms temporal resolution and D = 0 . 005 m2/s , which gives ≈ 5 cm displacement in 0 . 5 s [8] . We simulated the activity of N = [10 , 300] grid cells from a single module . Grid cells had a circular tuning curve: r k ( x ) = r max ( sin ( 2 π x 2 λ − ϕ k ) ) n + r 0 ( 13 ) with the following parameters: rmax = 15 Hz , r0 = 0 . 1 Hz , λ = 0 . 25 m and ϕk chosen to uniformly cover the interval [0 , 2π] . The power n = 22 was set to match the mean firing rate of the grid cells , 〈r ( x ) 〉 = 2 . 5 Hz , to experimental data [16] . Larger ( λ = 2 . 5 m ) grid spacing was modelled by decreasing the speed of the animal by a factor of 10 ( D = 0 . 00005 m2/s ) . The firing rate is shown in Fig 1b , right ( olive ) . Spike trains were generated as an inhomogeneous Poisson process with neurons conditionally independent given the simulated location: P ( s t k | x t ) = Poisson ( Δ t r k ( x t ) ) ( 14 ) Spikes of the neurons in module i , s0:t , i , represent the spatial location of the animal with error δ αi ( i . e . , with the same δ phase error for all modules ) which can be interpreted as the width of the ( periodic ) posterior probability distribution P ( x|s0:t , i ) . For an ideal observer this posterior distribution quantifies how much a given spatial location is consistent with the observed spike pattern . The posterior distribution of the position was numerically calculated by recursive Bayesian filtering: P ( x t | s 0 : t ) ∝ ∏ k P ( s t k | x t ) ∫ P ( x t - 1 | s 0 : t - 1 ) P ( x t | x t - 1 ) d x t - 1 ( 15 ) The colormap in Fig 1b shows this posterior distribution with N = 50 cells and λ = 0 . 25 m . Naturally , the width of the posterior depends on several factors , most importantly on the number of neurons observed in a given module and on the scale of the modules relative to the typical speed of the animal [8] . At each timestep the posterior distribution was fitted with a von Mises distribution with a location μt and a concentration parameter κt . The width of the posterior relative to the grid scale was estimated as: δ t = λ 2 π κ t ( 16 ) For analytic tractability , we use a bounded noise model in the derivations assuming that the location decoded from the spikes of a module is within δ αi distance from the true location . To be conservative , we chose δ to be the 99% of the empirical CDF of δt . The largest δ = 0 . 12 was found with λ = 0 . 25 m and N = 10 cells . The smallest δ = 0 . 01 corresponds to the parameters λ = 2 . 5 m and N = 300 cells . We assume that the modules are conditionally independent given the location of the animal , and hence position decoding , or representation , can be implemented by an ideal observer independently reading out the spikes , si , emitted by the different modules: P ( x|s ) = ∏i P ( x|si ) . When loosely talking about interference between the grid modules at a spatial point we refer to the interference between these periodic posterior distributions P ( x|si ) , i . e . , all module posteriors being larger than 0 at a location different from the origin ( Fig 1c ) . Since we measure the distance in units of the smallest grid scale ( α0 = 1 ) , avoiding interference at integer distances from the origin also guarantees the absence of interference elsewhere , i . e . , all positions in the interval [0 , L] will be distinguishable by the grid code . Hence we loosely call ϵ ( ℓ ) defined in Eq 1 the phase difference , but note that it is the phase difference at integer distance ℓ . Indeed , if the grid code was ambiguous confusing spatial locations x1 and x2 , then it would also confuse the origin with |x1 − x2| as well , since the phase differences of each module are the same between 0 and |x1 − x2| and between x1 and x2 ( Fig 2b and 2c , right ) . But |x1 − x2| can be confused with the origin only if |x1 − x2| is an integer , that is a multiple of the smallest scale , 1 . Note that this argument is correct only if the phase representation ambiguity of each module is independent of the actual position , which holds if we suppose that firing fields of cells from the same module are spaced evenly , which we do assume . Graphically , interference between locations occurs when two segments of the phase curve come close to each other . Since the segments of the phase curve are parallel ( Fig 2 ) , and we started the phase curve in the origin , interference first occurs in the origin . Avoiding interference at the origin as much as possible at arbitrary distances thus also guarantees that the segments of the phase curve are separated from each other as much as possible , leading to a uniform coverage of the phase space [14] . We call a real number α algebraic of order n ( positive integer ) , if n is the least integer such that α is the root of a polynomial of degree n with integer coefficients . Algebraic numbers of order one are exactly the rational numbers . Another example is the golden ratio , σ , which is irrational , and is the root of x2 − x − 1 , a integer polinomial of degree two . Therefore , σ is an algebraic number of order two . Since we fixed the resolution , the capacity of the code is proportional to the coding range . Moreover , as the coding precision of the modules was the same , we assume that the population size of each module is approximately N for grid scales chosen randomly from a bounded interval . The information rate of the grid system , defined as the ratio of the logarithm of the capacity and the total number of conveyed bits [14] is ρ ∝ 1 r ¯ N M log ( c α δ ) M ( 17 ) ∝ 1 r ¯ N log c α δ ( 18 ) ∝ 1 r ¯ N log c α log N k ( 19 ) where r ¯ is the average firing rate of a grid cell and in the third line we used that δ = k/log ( N ) [14] . Thus , the information rate is independent of the number of modules and increases with log cα . For a geometric code with scale ratio α the optimal population size for dynamical decoding and constant δ decreases as n i = n 0 / λ i 2 = n 0 / α 2 i where λi = αi is the scale of module i and n0 is the number of neurons in the first module [8] . In this case the total number of neurons in the population is N = ∑ i = 0 M n 0 / α 2 i = n 0 α 2 α 2 - 1 ( 20 ) Since the total number of neurons does not grow linearly with the number of modules , the information rate becomes proportional to M: ρ ∝ M α 2 - 1 n 0 α 2 r ¯ log c α δ ( 21 ) Although a constraint on the minimal number of cells per module will limit the finite information rate to remain finite , Eq 21 emphasises that adding further modules with larger periods increases the efficiency of the grid system if the number of cells per module is set optimally for dynamical decoding [8] . Although a geometric progression of scales is consistent with both nested and MA codes , the information rate is higher for optimal nested codes since they maximise α . In this section we demonstrate that a set of grid cells with scale ratio ( α ) optimally chosen between pairs of successive grid modules is close to being pessimal for efficient space representation for M > 2 . Such pairwise optimisation leads to a set of scales showing geometric progression with the scale ratio being α , i . e . , [1 , α , α2 , …] , which is consistent with the experimental data [10 , 12 , 23 , 13] . The representation of the position becomes ambiguous if all modules show interference at the same location , i . e . , the phase of all modules are very close to 0 at distance ℓ from the origin . Consider for example the golden ratio α = σ , which is a second order algebraic number , i . e . , it is the root of the integer coefficient polynomial x2 − x − 1 . Therefore , the phase ψ2 ( x ) = ( x mod σ2 ) /σ2 of any spatial point x according to the third module can be simply expressed with the phase of the first two modules as ψ 2 ( x ) = [ ψ 0 ( x ) - ψ 1 ( x ) ] mod 1 . ( 22 ) To see this , consider that by the definition of the phases ψi ( x ) when the animal is at distance x from the origin there are some integers ℓ , k1 , k2 so that x = ℓ + ψ 0 ( x ) = σ ( k 1 + ψ 1 ( x ) ) = σ 2 ( k 2 + ψ 2 ( x ) ) . Using that σ2 − σ − 1 = 0 we get that σ 2 ( ℓ + ψ 0 ( x ) ) - σ 2 ( k 1 + ψ 1 ( x ) ) - σ 2 ( k 2 + ψ 2 ( x ) ) = 0 . Rearranging terms yields ψ 2 ( x ) = ℓ - k 1 - k 2 + ψ 0 ( x ) - ψ 1 ( x ) = [ ψ 0 ( x ) - ψ 1 ( x ) ] mod 1 . In other words , the phase of the third module provides no additional information given the phase of the other two modules . In particular , if both ψ0 ( x ) and ψ1 ( x ) are close to 0 ( Fig 6b ) , then so is ψ2 ( x ) and hence the third module fails to resolve the ambiguity when the two first modules interfere . Similarly , if we have n grid cell modules with scales 1 , α , … , αn−1 with α being an algebraic number of order k < n , then all of the n phases can be expressed by any k of them , leading to redundant and inefficient representation . Clearly the same argument works not only for the powers of the golden ratio , but for powers of any algebraic number of order lower than the number of modules . To derive the general solution for M grid modules , we consider a set of 1-dimensional grids with scales α0 = 1 < α1 < ⋯ < αM −1 . Again , the interference between the modules can be expressed by the simultaneous Diophantine approximation of the vector A = ( α 1 , … , α M - 1 ) using fractions of integers with the common numerator ℓ , i . e . , αi ≈ ℓ/ki . Importantly , a theorem by Dirichlet provides an upper bound on the efficiency of the approximation . Namely , for all ( M − 1 ) -tuple of irrational numbers α1 , … , αM −1 we have infinitely many collections of integers k0 , k1 , … , kM −1 ( with k0 = ℓ ) , such that the approximation error defined as ϵ ˜ i j ( l ) = | k i α i − k j α j | ( 23 ) is simultaneously smaller than the upper bound for all items in the tuple: ϵ ˜ i j ( ℓ ) < α i + α j ℓ 1 / ( M - 1 ) ( ∀ i , j = 0 , … , M - 1 , i ≠ j ) . ( 24 ) Note , that ϵ ˜ i j differs from ϵ defined for two modules ( Eq 2 ) as it is not normalised with α . Proof of Eq 24 . First we prove that any vector of irrationals can be approximated to the claimed order with rationals having the same denominator . Let A = ( α 1 , … , α n - 1 ) . To approximate A with rationals of denominator at most Q let us define the vectors a j = j A - ⌊ j A ⌋ , j = 0 , … , Q , where floor is understood coordinate-wise . Let us partition the unit cube [0 , 1]n−1 into small cubes of side length Q−1/ ( n−1 ) , so that altogether we have Q of them . Since we have Q + 1 many aj-s each falling into [0 , 1]n−1 , hence there will be ( at least ) 2 of them falling into the same small cube , ak and al , say . Then | | k - l | A - | ⌊ k A ⌋ - ⌊ l A ⌋ | | ≤ | a k - a l | ≤ Q - 1 / ( n - 1 ) , with the inequalities holding coordinate-wise . Therefore , because of |k − l| ≤ Q , A is approximable with denominator |k − l| and numerator ( vector ) | ⌊ k A ⌋ - ⌊ l A ⌋ | with error not exceeding |k − l|− ( 1+1/ ( n−1 ) ) . The desired statement follows then by simultaneously approximating the numbers 1/αi with common denominator , which is also a simultaneous approximation of αi with common numerator , which completes the proof . For a set of grid scales αi = αi ( i = 0 , … , M − 1 ) where α is an algebraic number of degree M , there exists a maximal positive constant c A , such that ϵ ˜ ( ℓ ) = max i , j { 1 α i + α j ϵ ˜ i j ( ℓ ) } > c A ℓ 1 / ( M - 1 ) ( 25 ) holds , except for at most finitely many integers ℓ . To see that Eq 25 holds , we start from the work of [44] ( see also [45] ) stating that powers of an algebraic number are badly simultaneously approximable with common denominator in the following sense . Let β be an algebraic number of order M . There exists cβ > 0 such that for all integer ℓ , ki there is i ∈ {1 , … , M − 1} for which | β i ℓ - k i | > c β ℓ 1 / ( M - 1 ) . Derivation of Eq 25 . Our goal is to give a lower bound on |αi ki − αj kj| , where α is algebraic of order M , 0 ≤ i , j ≤ M − 1 . Without loss of generality suppose that i < j . | α i k i - α j k j | = | α i - j k i - k j | α j > α j c A k i 1 / ( M - 1 ) . Now the fact that ki ∼ ℓ/αi implies Eq 25 if c A > 0 is chosen appropriately . The position representation is unambiguous if there is at least one pair of modules for which the phase difference is larger than the threshold set by the noise , i . e . , ϵ ˜ i , j ( ℓ ) > δ ( α i + α j ) which holds if δ < c A ℓ 1 / ( M - 1 ) ( 26 ) From here , the critical distance Lmax up to which coding is unambiguous can be expressed as ( cf . Eq 9 ) : L max ≔ ( c A δ ) M − 1 , ( 27 ) for all δ which is small enough . To directly compare the capacity of the MA grid cell system derived in Eq 9 with previous estimates for nested coding [4 , 6] , we also calculate Nmax , the number of distinguishable spatial phases: N max ≔ L max 2 δ = 1 2 c A ( c A δ ) M ( 28 ) Efficient coding with nested modules requires that αi = ri with 0 ≤ i ≤ M − 1 and r being the scale ratio with fixed relative uncertainty of modules 2δ = 1/r [6] . The position of the animal can be determined at precision 1/r without ambiguity if the animal is restricted to move in an environment with the size identical to the scale of the largest module , rM−1 . In this case the number of distinguishable spatial phases is r M = ( 1 / 2 δ ) M , which is identical to the capacity we found for non-nested coding when c A = 0 . 5 ( Eq 28 ) . To derive Eq 27 we first show that interference of the grid representation is equivalent to pairwise interference between all pairs of modules . To test unambiguity of coding note that the place at distance x from the origin is confusable with 0 if for all i = 0 , … , M − 1 there exists an integer ki such that | k i α i − x | < α i δ , ( 29 ) where δ is the relative uncertainty of modules . It turns out that , as for M = 2 , there is no need to consider all x ∈ [0 , Lmax] , it is enough to care with integers: Claim . There exists x ∈ [0 , Lmax] for which Eq 29 holds for all i exactly when the following pairwise interference occurs between all modules: | k i α i - k j α j | < ( α i + α j ) δ ( 30 ) for all i , j with some integers ki ( i = 0 , … , M − 1 ) such that 0 < ki αi ≤ Lmax . Proof . Let us fix ki , i = 0 , … , M − 1 . Pairwise interference means that there is a point xi , j in the intersection of ( ki αi − αi δ , ki αi + αi δ ) = ( ai , bi ) and ( kj αj − αj δ , kj αj + αj δ ) = ( aj , bj ) . Due to the topology of the line , it is easy to see by induction that the intersection of all such intervals is nonempty and hence one can chose xi , j = x . The statement is obvious for M = 2 . Now suppose that the intersection ∩ i = 0 n ( a i , b i ) ≠ ∅ . Then it is the interval ( a , b ) with a = max i = 0 , … , n a i and b = min i = 0 , … , n b i . If ( an+1 , bn+1 ) intersects ( ai , bi ) , then both an+1 < bi and bn+1 > ai , and therefore an+1 < b and bn+1 > a , which completes the induction . Therefore Eq 29 implies Eq 30 . The other direction is immediate . Now using the above Claim Equation Eq 27 easily follows by rearranging Eq 25 . Let us fix the relative uncertainty of modules δ < 1/2 and a number αmax > ( 1 + δ ) / ( 1 − δ ) . We show that if scales α1 , α2 , … are drawn uniformly at random from [1 , αmax] , independently of each other , then for any δ < ζ < 1/2 the representation with M modules having scales α1 , α2 , … , αM is unambiguous in every spatial position x > 0 up to X max ≔ ( ζ δ ) M − 1 ( 31 ) with probability of order 1 − ( 2ζ ) M as M → ∞ . Here ζ is the analog of c A which characterises the capacity of a particular grid cell system . As we will see , the convergence holds for any ζ < 1/2 , but the speed of the convergence depends on ζ: higher efficiency is guaranteed to be achieved only for larger number of modules . Proof: Let α1 , α2 , … be independent random variables distributed uniformly on [1 , αmax] . Let x be a spatial point and let ψ ˜ i = ψ ˜ i ( x , α i ) denote the phase of module i ( with scale αi ) at x , that is ψ ˜ i ≔ ( x mod α i ) / α i = x / α i mod 1 = x / α i - ⌊ x α i ⌋ ∈ [ 0 , 1 ] . Note that for fixed x the distribution of phases ψ ˜ i are independent of each other since the α-s are independent . We also use the notation p1 ( x ) for the probability that the phase ψ ˜ i is ( almost ) indistinguishable from 0 , defined in the following way: p 1 ( x ) = P α i ( ψ ˜ i ∈ [ 0 , ( 1 + ε ) δ ] ∪ [ ( 1 - ( 1 + ε ) δ ) , 1 ] | x ] . where ε > 0 is determined later . It is easy to see that p1 ( x ) does not depend on i , i . e . , it is the same for all modules . Moreover , the distribution of ψ ˜ i converges to uniform as the distance increases , in particular limx → ∞ p1 ( x ) = 2 ( 1 + ε ) δ . The convergence of this distribution to the uniform is a key observation that remains true even in higher dimensions with uniform random rotations or in case of slight variation of the grid scales on the long range . Hence there exists a critical distance , x0 = x0 ( δ , ε ) for which all x > x0 we have |p1 ( x ) − 2 ( 1 + ε ) δ| ≤ δε . Therefore , for x > x0 we have p 1 ( x ) ≤ ( 2 + 3 ε ) δ . ( 32 ) It also implies a bound on the probability of interference of many modules at a given point x . If we consider M modules with scales drawn uniformly at random from [1 , αmax] and independently of each other , then by Eq 32 for x > x0 the probability of all phases being close to 0 is p M ( x ) = ℙ ( ( ∀ i ≤ M ) ψ ˜ i ∈ ( 0 , ( 1 + ε ) δ ) ∪ ( ( 1 − ( 1 + ε ) δ ) , 1 ) | x ) = p 1 ( x ) M ≤ ( ( 2 + 3 ε ) δ ) M , ( 33 ) that is , pM ( x ) is exponentially small in M . There remains to estimate the probability of interference of many modules anywhere up to a maximally allowed spatial distance . Our goal is to show that P ( ( ∃ x < X max ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , δ ) ∪ ( ( 1 - δ ) , 1 ) ) → 0 ( 34 ) as M → ∞ , where X max = ( ζ δ ) M − 1 , as in Eq 31 . Note , that satisfying Eq 34 is not trivial , since Xmax increases exponentially with M . There is no need to investigate all x < Xmax , it is enough to show , that there is no interference on a set which is dense enough in [0 , Xmax] in the stronger sense of Eq 33 . Indeed , let Y be an ε dense set in [0 , Xmax] with at most 2Xmax/ε elements . Then { ( ∃ x < X max ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , δ ) ∪ ( ( 1 − δ ) , 1 ) } ⇒ { ( ∃ x ∈ Y ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , ( 1 + ε ) δ ) ∪ ( ( 1 − ( 1 + ε ) δ ) , 1 ) } where we used the fact that the d d x ψ ˜ i = 1 α i < 1 since αi was chosen from the interval [1 , αmax] . The corresponding inequality for the probabilities of these events is ℙ ( ( ∃ x < X max ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , δ ) ∪ ( ( 1 − δ ) , 1 ) ) ≤ ℙ ( ( ∃ x ∈ Y ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , ( 1 + ε ) δ ) ∪ ( ( 1 − ( 1 + ε ) δ ) , 1 ) ) . Now for these finitely many points x ∈ Y we can use Eq 33 one by one , if x > x0: ℙ ( ( ∃ x 0 < x ∈ Y ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , ( 1 + ε ) δ ) ∪ ( ( 1 − ( 1 + ε ) δ ) , 1 ) ) ≤ ( ( 2 + 3 ε ) δ ) M 2 X max / ε = 2 ε ( ( 2 + 3 ε ) δ ) M ( ζ δ ) M − 1 < 1 ε ζ ( ζ ( 2 + 3 ε ) ) M → 0 if ε < 1 - 2 ζ 3 ζ , which we assume , where in the first inequality we used Eq 33 and union bound , and then in the second one that X max = ( ζ δ ) M − 1 . We have to remark that interference in different spatial points is not independent of each other , but union bound works even in that case . There remains to show that the grid cell representation works up to x0 . Clearly there is no ambiguity up to x = 1 + δ . To estimate the probability P ( ( ∃ x 0 ≥ x ∈ Y ) ( ∀ i ≤ M ) ψ ˜ i ( x ) ∈ ( 0 , ( 1 + ε ) δ ) ∪ ( ( 1 - ( 1 + ε ) δ ) , 1 ) ) ( 35 ) we first have to observe that the cardinality of Y ∩ [1 + δ , x0] is independent of M . Therefore to guarantee that the probability in Eq 35 goes to 0 we need to show that for all 1 + δ ≤ x ≤ x0 there is a scale α ∈ [1 , A] which is able to distinguish x from the origin , that is α such that ψ ˜ ( α ) = ( x / α mod 1 ) ∈ [ δ , 1 - δ ] . This is so because x/α is monotonically decreasing in α and because x / 1 - x / A ≥ x ( 2 δ / ( 1 + δ ) ) ≥ 2 δ , where we used that αmax > ( 1 + δ ) / ( 1 − δ ) and x > 1 + δ . Therefore ψ ˜ ( α ) can not lay in [0 , δ] ∪ [1 − δ , 1] for all α ∈ [1 , αmax] . The constant c A ( and cα ) is well defined only for algebraic numbers , but can also be estimated for real numbers from the scaling of the phase difference with distance using numerical simulations . As c A is defined asymptotically ( Eq 25 ) , in order to estimate it numerically we need an approximation of it for finite distances . An alternative definition of c A ( equivalent with Eq 25 ) is c A = lim inf ℓ → ∞ ϵ ^ A ( ℓ ) , ( 36 ) where ϵ ^ A ( ℓ ) is defined by ϵ ^ A ( ℓ ) = min K , k 0 = ℓ max i , j { | α i k i - α j k j | / ( α i + α j ) } ℓ 1 / ( M - 1 ) , ( 37 ) where K = ( k 1 , … , k M - 1 ) . Intuitively , to find the magnitude of interference at location ℓ , for all possible values of K we first select the maximum phase difference in the set and then choose the set with the smallest maximum . From the plots Figs 4 and 7 it is clear that the naive way of approximating c A with c A ( ℓ ) for some large ℓ is not a good idea , as c A ( ℓ ) may vary heavily with ℓ , especially for non-algebraic scale ratios . Note , that the calculation of c ^ α is a special case of c ^ A with M = 2 . To estimate coding efficiency in the presence of noise we are mostly interested in the above infemum when ℓ is such that the phase difference ϵ ^ A ( ℓ ) / ℓ 1 / ( M - 1 ) is close to the precision δ of the modules . It motivates to investigate the ( numerically computable ) minimum c ^ A ( δ 1 , δ 2 ) = min { ϵ ^ A ( ℓ ) | ℓ 2 ( δ 2 ) ≤ ℓ ≤ ℓ 1 ( δ 1 ) } for some pair δ1 < δ2 , where ℓ2 is so that for all ℓ ≥ ℓ2 we have ϵ ^ A ( ℓ ) / ℓ 1 / ( M - 1 ) < δ 2 and ℓ1 is the smallest ℓ so that ϵ ^ A ( ℓ ) / ℓ 1 / ( M - 1 ) < δ 1 . A common and natural way to numerically investigate Diophantine approximation is using lattice reduction [46] . By lattice we mean a subset L of R d defined by some vectors v 1 , … , v m ∈ R d , m ≤ d so that L = { w = ∑ i = 1 m b i v i ∣ b i ∈ Z } . Given a lattice L , a classical computational problem is to find the shortest non-zero vector of it ( Fig 8 ) . In the followings we show how Diophantine approximation of a vector ( α1 , … , αn ) can be investigated with the help of finding shortest vectors of appropriately chosen lattices . Let us first consider a simple example . Let the lattice L be defined by the rows of the matrix V = [ v 1 ⋮ v n + 1 ] ≔ [ α 1 0 … 0 0 0 α 2 … 0 0 ⋮ ⋮ ⋱ ⋮ ⋮ 0 0 … α n 0 - 1 - 1 … - 1 ε ] , where ε > 0 . For all ε which is small enough the shortest vector w 0 = ∑ i = 1 n + 1 b i v i of L corresponds to a simultaneous Diophantine approximation of ( α1 , … , αn ) with the common numerator bn+1 and denominators bi , i = 1 , … , n . The parameter ε can be considered as a penalty term: the smaller this term the bigger the numerator can be . When speaking about shortest vectors we need to specify the norm with respect to which vectors are compared . Here we are looking for the largest phase difference between the modules so we use supremum norm ( Eq 25 ) . The shortest vector in supremum norm of the lattice defined by V is an approximation so that max { b n + 1 ε , max i { | b n + 1 - b i α i | } is as small as possible . By this we can compute what is the maximal phase difference between the module with scale 1 and all other modules up to distance bn+1 . Remember that according to Eq 25 we are searching for an approximation minimizing max { b n + 1 ε , max i , j { | α i k i - α j k j | / ( α i + α j ) } ℓ 1 / ( M - 1 ) } . ( 38 ) Similarly to the previous example , it can be done simply by dividing columns i , i = 1 , … , n of V by ( 1 + αi ) , and by adding some more columns of similar form which refer to interference between modules i and j . For example , for n = 3 the shortest ( in sup norm ) element of the lattice generated by the rows of the following matrix gives an approximation minimizing Eq 38: [ α 1 1 + α 1 0 0 α 1 / α 3 α 1 + α 3 0 α 1 / α 2 α 1 + α 2 0 0 α 2 1 + α 2 0 0 α 2 / α 3 α 2 + α 3 - 1 α 1 + α 2 0 0 0 α 3 1 + α 3 - 1 α 1 + α 3 - 1 α 2 + α 3 0 0 - 1 1 + α 1 - 1 1 + α 2 - 1 1 + α 3 0 0 0 ε ] . In this way maximal interference in the grid cell system can be computed numerically as shortest vectors of some lattices in supremum norm . Finding this shortest vector is an integer linear programming ( ILP ) problem , which in general is an NP-hard computational problem , and can be solved by e . g . a branch and bound algorithm [47] . There are also efficient methods which find approximation solutions in polynomial time , such as the LLL algorithm due to Lenstra , Lenstra and Lovász [46] . The LLL algorithm finds not only a short vector of a lattice , but also another basis of it which consists of short and nearly orthogonal vectors in the L2 norm , a so called LLL reduced basis . The error made by the LLL algorithm is too high to precisely compute the constant terms in Eq 27 , and therefore we could not rely only on this algorithm . Nevertheless , compared to the ILP solution , we could significantly speed up our computations by first applying the LLL algorithm to find an approximate solution ( and a reduced lattice ) , and then an ILP solver on this LLL reduced basis , which could find nontrivial optimal solutions very efficiently if started from this input .
|
Navigation in natural , open environments poses serious challenges to animals as the distances to be represented may span several orders of magnitudes and are potentially unbounded . The recently discovered grid cells in the rodent brain are though to play a crucial role in generating unique representations for a large number of spatial locations . However , it is unknown how to choose the parameters of the grid cells to achieve maximal capacity , i . e . , to uniquely encode the utmost locations in an open environment . In our manuscript , we demonstrate the surprising robustness of the grid cell coding system: The population code realised by grid cells is close to optimal for unique space representation irrespective of the choices of grid parameters . Thus , our study reveals a remarkable robustness of the grid cell coding scheme and provides a solid theoretical explanation for the large diversity of the grid scales observed in experimental studies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"decision",
"making",
"brain",
"social",
"sciences",
"neuroscience",
"number",
"theory",
"optimization",
"cognitive",
"psychology",
"mathematics",
"cognition",
"computational",
"neuroscience",
"real",
"numbers",
"neuronal",
"tuning",
"coding",
"mechanisms",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"cellular",
"neuroscience",
"psychology",
"entorhinal",
"cortex",
"cell",
"biology",
"anatomy",
"neurons",
"information",
"theory",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"computational",
"biology",
"cognitive",
"science",
"cerebral",
"cortex"
] |
2018
|
Robust and efficient coding with grid cells
|
Multicellular animals match costly activities , such as growth and reproduction , to the environment through nutrient-sensing pathways . The insulin/IGF signaling ( IIS ) pathway plays key roles in growth , metabolism , stress resistance , reproduction , and longevity in diverse organisms including mammals . Invertebrate genomes often contain multiple genes encoding insulin-like ligands , including seven Drosophila insulin-like peptides ( DILPs ) . We investigated the evolution , diversification , redundancy , and functions of the DILPs , combining evolutionary analysis , based on the completed genome sequences of 12 Drosophila species , and functional analysis , based on newly-generated knock-out mutations for all 7 dilp genes in D . melanogaster . Diversification of the 7 DILPs preceded diversification of Drosophila species , with stable gene diversification and family membership , suggesting stabilising selection for gene function . Gene knock-outs demonstrated both synergy and compensation of expression between different DILPs , notably with DILP3 required for normal expression of DILPs 2 and 5 in brain neurosecretory cells and expression of DILP6 in the fat body compensating for loss of brain DILPs . Loss of DILP2 increased lifespan and loss of DILP6 reduced growth , while loss of DILP7 did not affect fertility , contrary to its proposed role as a Drosophila relaxin . Importantly , loss of DILPs produced in the brain greatly extended lifespan but only in the presence of the endosymbiontic bacterium Wolbachia , demonstrating a specific interaction between IIS and Wolbachia in lifespan regulation . Furthermore , loss of brain DILPs blocked the responses of lifespan and fecundity to dietary restriction ( DR ) and the DR response of these mutants suggests that IIS extends lifespan through mechanisms that both overlap with those of DR and through additional mechanisms that are independent of those at work in DR . Evolutionary conservation has thus been accompanied by synergy , redundancy , and functional differentiation between DILPs , and these features may themselves be of evolutionary advantage .
The ability of organisms to respond appropriately to changes in their environment is key to survival and reproductive success . An essential environmental variable for all organisms is their food supply and energetically demanding processes , such as growth , metabolism and reproduction are matched to nutrition by nutrient-sensing pathways , such as the insulin/IGF signalling ( IIS ) and TOR pathways [1] . An important recent discovery has been that reduced activity of IIS and TOR can slow aging and increase stress resistance and lifespan in the yeast Saccharomyces cerevisiae , the nematode worm Caenorhabditis elegans , the fruit fly Drosophila melanogaster and mice [2] . The mechanisms by which these pathways exert their diverse effects are hence of interest , as are the ways in which these parallel biological roles are achieved in evolutionarily diverse organisms . The IIS pathway includes both peptide ligands , which can act at a distance , and intracellular components . In mammals , the ligands include insulin , the insulin-like growth factors ( IGF ) and relaxins . IGFs are mainly involved in growth control during development , whereas insulin secretion from pancreatic β-cells controls carbohydrate and lipid metabolism . Relaxins are produced by the ovary and are involved in reproduction . Insulin-like peptides ( ILPs ) have also been identified across a broad range of invertebrates , including molluscs , the nematode Caenorhabditis elegans and several insect species [3] . Most invertebrate genomes contain multiple ILPs , including 40 in C . elegans [4] and 7 in Drosophila melanogaster ( DILP1-7 ) [5] . In contrast , while mammals often have up to 4 isoforms of the cellular components of IIS , they are encoded by single genes in Drosophila , including one Drosophila Insulin receptor ( DInR ) , one insulin receptor substrate ( chico ) and one downstream forkhead box O transcription factor ( dFOXO ) [1] . The relative simplicity of the cellular IIS pathway , together with the diversification of DILPs , implies that the diverse functions of IIS could be in part mediated by functional diversification of the ligands . Supporting functional differentiation between ligands , each dilp gene shows a characteristic spatio-temporal expression pattern . For instance , DILP4 is expressed in the embryonic midgut and mesoderm [5] , DILP6 predominantly in the larval and adult fat body , with expression strongly up-regulated in the transition from larva to pupa [6] . DILP7 is expressed in specific neurons that innervate the female reproductive tract [7] , [8] , and inactivation of them results in sterile flies with an “egg-jamming” phenotype , suggesting that DILP7 could be a Drosophila relaxin [7] . DILP1 , 2 , 3 and 5 are expressed in brain median neurosecretory cells ( MNCs ) of the larval brain [5] , [9] , [10] , but only DILP2 , 3 and 5 could be detected in MNCs of the adult fly [11] . DILP2 is also expressed during development in imaginal discs , and in salivary glands and DILP5 in follicle cells of the female ovary [5] . Targeted ablation of the MNCs during early larval development results in developmental delay , growth defects and elevated carbohydrate levels in the larval hemolymph [10] , while later ablation , during the final larval stage , results in lower female fecundity , increased storage of lipids and carbohydrates , elevated resistance to starvation and oxidative stress and increased lifespan [11] . Notably , expression of ILPs in MNC is evolutionarily conserved among insects , suggesting important evolutionarily conserved functions for these cells and their ligands . Furthermore , similar developmental programs are involved in the specification of MNCs in Drosophila and pancreatic beta cells in mammals , suggesting that insulin-producing cells of invertebrates and vertebrates may be derived from a common ancestry [12] . It is not yet clear if the phenotypes of MNC-ablated flies result from loss of one or more of the DILPs , or whether MNCs have functions independent of DILPs . Nor is it known if MNC-expressed DILPs have specific functions or act redundantly . All seven DILPs possess the ability to promote growth , with DILP2 the most potent , and these ligands therefore probably all act as DInR agonists [9] . Over-expression of DILP2 also suppressed germ line stem cell loss in ageing females , probably through action in the stem cell niche [13] . DILP2 may also modulate lifespan , because its transcript is lowered in various mutant , long-lived flies . Over-expression of dFOXO in the adult fat body [14] , over-expression of a dominant negative form of p53 in MNCs [15] , increased JNK activity in MNCs [16] and hypomorphic mutants of the Drosophila NPY like protein sNPF [17] , have all been reported to both extend lifespan and reduce the level of DILP2 expression . However , direct reduction in the level of DILP2 by in vivo double-stranded RNA interference ( RNAi ) did not increase lifespan [18] , leaving the role of DILP2 unclear . Dietary restriction ( DR ) , a reduction in food intake without malnutrition , extends lifespan in diverse animals . In both C . elegans and Drosophila extension of lifespan by DR has been suggested to be independent of IIS , because lifespan increases in response to DR in animals lacking the key IIS effector FOXO transcription factor [19]–[21] . However , the finding could indicate instead that , in the absence of a normal increase in FOXO activity during DR , other pathways can act redundantly to increase lifespan . Indeed , other evidence has suggested that reduced IIS and DR may extend lifespan through overlapping mechanisms [19] , [22] . DILP3 and DILP5 transcript levels are reduced in starved larvae [9] and DILP5 in DR adult flies [20] , potentially indicating a role of DILPs in the fly's response to DR . Gene families generally expand by gene duplication , and the duplicate copies are retained either if there is a requirement for large amounts of the gene product or if the duplicate copies undergo some sequence divergence and functional differentiation [23] . Recent work has suggested that feedback between partially redundant duplicate genes could itself be an important source of information aiding signal transduction [24] . It is not known if , as well as the functional differentiation implied by the findings described above , there is functional redundancy between the Drosophila ILPs . Nor is it known if there is stability of sequence differentiation and of membership of this gene family over evolutionary time or whether there is gene turnover . In addition , assignment of specific functions to individual dilp genes requires experimental manipulation . Although studies using RNAi have been informative [17] , [18] , RNAi can be prone to off-target effects [25] and other forms of cellular toxicity , and often results only in hypomorphic phenotypes . We have used the completed genome sequences of 12 Drosophila species [26] to examine the stability of the dilp gene family , and found that the 7 DILPs have remained present and clearly differentiated from each other in sequence over a period of 40–60 million years . Evolutionary conservation of different regions of the peptides suggests that 6 of the 7 DILPs are cleaved like mammalian insulins , while the seventh may remain uncleaved , like mammalian IGFs . We generated specific null mutants for all seven DILPs , by homologous recombination or P-element mediated excision , and also generated flies that lack two or more DILPs simultaneously . Using these mutants we made a systematic analysis of DILP function in development , metabolism , reproduction , stress resistance , lifespan and response to DR of Drosophila . We show that DILPs can act redundantly , which suggests that redundancy among ILPs may be of evolutionarily advantage . We found both synergy and compensation of expression between DILPs . In particular DILP6 in the fat body compensated for the loss of MNC DILPs , demonstrating that DILPs are part of a complex feedback system between the central nervous system and peripheral tissues such as the fat body , which controls development , metabolism and reproduction . We further show that DILP2 is an important determinant of lifespan , describe a novel role for the fat body derived DILP6 peptide in growth control and demonstrate that dilp7 null mutants have normal fecundity , contrary to the suggestion DILP7 could be a Drosophila relaxin . Finally , we describe a specific interaction between the endosymbiont Wolbachia and IIS in the regulation of lifespan and show that DILPs mediate the responses of lifespan and fecundity to DR in Drosophila .
Taking advantage of the 12 sequenced Drosophila genomes [26] , we investigated the evolution of ILPs across the Drosophila genus . As in D . melanogaster , seven dilp genes were identified in all 12 Drosophila species , with the exception of D . simulans and D . grimshawi ( Figure 1 ) . Using the Jones-Thornton-Taylor ( JTT ) matrix to calculate average amino acid distances , we found that the distances between the seven DILPs within each species is much greater than the distances among the putative orthologues in the different Drosophila species ( Figure 1 ) We did not identify a dilp6 gene in D . simulans , probably because of sequencing gaps in the corresponding genomic region . Thus the seven dilp genes were present before the divergence of the 12 Drosophila species more than 40 million years ago , and none has been turned over during that period . Interestingly , the D . grimshawi genome contains eight dilp genes , as a result of a duplication of dilp2 ( Figure 1 ) . D . grimshawi is a member of the endemic Hawaiian Drosophila species , which are characterized by their large body size . The finding that DILP2 is an important regulator of growth in D . melanogaster ( [9] and see below ) suggests that evolutionary changes in dilp2 gene expression may have contributed to the evolution of body size in the Hawaiian Drosophila species . In mammals , IIS ligands are synthesized as pre-propeptides , consisting of a signal peptide and contiguous B-C-A peptides . In insulin and relaxin the C-peptide is clipped out by a convertase enzyme targeting basic amino acid cleavage sites , to produce a bioactive peptide consisting of A- and B-chain linked by 2–3 disulfide bridges . In contrast , IGFs contain a shortened C-peptide , which is not removed , resulting in a single chain peptide hormone . We therefore examined evolutionary conservation of different regions of the DILPs across the 12 Drosophila species to determine if the proteins are likely to be cleaved . Amino acid alignments of DILP pre-propeptides from the different Drosophila species shows that the bioactive peptides , i . e . the A and B chains are more conserved than the signal peptides and the C-peptides ( Figure S1 ) . The only exception is the shortened C-peptide in DILP6 , which shows a similar degree of conservation to the A and B chains . This finding may indicate that although it contains a basic cleavage site the C-peptide is not removed and is part of a single chain bioactive DILP6 peptide , as has been recently suggested for the DILP6-like BIGFLP protein in the silkworm Bombyx mori [6] . Thus , DILP6 may resemble IGF rather than insulin . Despite the low amino acid conservation , functional signal peptides were found to be present in all DILP pre-propeptides using the Signal P prediction software [27] , which indicates that all seven DILPs act as secreted peptide hormones in all 12 Drosophila species . Furthermore , cysteine residues involved in disulfide bridge formation as well as basic cleavage sites are highly conserved ( Figure S1A ) , further supporting the view that DILPs , except for DILP6 , resemble insulin and consist of heterodimeric peptides of A and B-chain linked by disulfide bridges . Although the seven DILPs have been stably differentiated and retained in the 12 Drosophila genomes , they show different degrees of amino acid conservation ( Figure S1B ) , suggesting different degrees either of functional constraint or of directional selection . DILP7 is by far the most highly conserved DILP peptide , with an overall amino acid identity of 76% between the pre-propeptides of D . melanogaster and the most distantly related Drosophila species D . grimshawi , increasing to 83% and 94% , respectively , when only the A and B chains are considered ( Figure S1B ) . Indeed , only DILP7 has bona fide orthologues outside the Drosophila family , with 64–66% amino acid identity ( A and B chain , 48% overall ) with ILP5 from the mosquitoes Anopheles gambiae and Aedes aegypti and 63% and 55% ( A and B chain , 48% overall ) with ILP7 from the red flour beetle Tribolium castaneum [28] , [29] . DILP4 is next most conserved , with only 52% overall amino acid identity and 75% and 70% identity in the A and B chain , respectively ( D . melanogaster vs . D . grimshawi ) . DILPs 1 , 2 , 3 , 5 and 6 evolved faster than DILP4 and DILP7 and at about the same rate as each other ( ca . 40% overall amino acid identities , 50–55% identities within the A and B chain , D . melanogaster vs . D . grimshawi , Figure S1B ) . The higher amino acid sequence conservation of DILPs 7 and 4 suggests that these two DILPs might carry essential functions that are different from the other DILPs and can therefore not be compensated by the other DILPs . In order to analyse the in vivo function of individual dilp genes , we generated dilp-specific mutants by ends-out homologous recombination [30] for dilp1 , 2 , 3 , 4 , 5 and 7 ( Figure 2A–2C ) and by P-element mediated imprecise excision for dilp6 ( Figure 2D ) . In addition , to address redundancy and synergy among individual dilp genes , we used homologous recombination to generate mutant flies with a combined knock out of two or several DILPs , including 2 , 3 ( dilp2–3 ) , 2 , 3 , 5 ( dilp2–3 , 5 ) , 1 , 2 , 3 , 4 ( dilp1-4 ) and 1 , 2 , 3 , 4 , 5 ( dilp1–4 , 5 ) . Ends out homologous recombination donor constructs were designed to delete the complete coding sequence of the targeted dilp gene , by replacing it with a white marker gene without affecting the genomic sequence of adjacent genes ( Figure 2A–2C ) . Putative homologous recombination events were tested by PCR on genomic DNA with gene-specific primer combinations for the insertion of the white marker gene into the target gene location ( Figure 2E ) . Several independent targeting events were recovered per dilp gene . Long-range PCR analysis on genomic DNA showed that most targeting events were precise homologous recombinations ( Figure S2 and data not shown ) and only these were used for subsequent experiments . Reverse Transcription ( RT ) PCR analysis showed that the dilp mutants are transcript-null alleles ( Figure 2F ) . Lack of DILP expression in the mutants was confirmed by immunohistochemistry on adult fly brains ( Figure S3 ) and by Western blot analysis ( Figure 3D ) . The single mutants were specific for individual dilp genes . For instance , dilp2 mutants lacked expression only of DILP2 , not of DILP3 or DILP5 ( Figure S3D , S3E , S3F ) . dilp6 mutant flies were generated by imprecise P-element excision of KG04972 integrated in the 5′upstream region of the dilp6 gene ( Figure 2D ) . Two different dilp6 alleles were isolated , the small deletion dilp641 covering the dilp6 5′upstream region including the first exon and the large deletion dilp668 covering the complete dilp6 gene as well as at least 4 other genes . RT PCR analysis confirmed dilp668 to be a transcript null allele ( Figure 2F ) . dilp6 expression was still detectable in dilp641 mutants . However , 5′ RACE analysis revealed that these transcripts lacked the first exon of the dilp6 gene and instead contained ectopic sequence including additional ORFs immediately upstream of the dilp6 ORF , which might interfere with the translation of the DILP6 peptide ( for details see Text S1 ) . Consistently , whenever tested both dilp6 mutant alleles were phenotypically indistinguishable from each other . In addition , dilp641 mutant flies showed the same developmental growth defects as flies in which dilp6 is knocked down by RNAi [31] , suggesting dilp641 to be a strong hypomorph or even a functional null allele . Reduced expression of DILPs has been associated with changes in IIS pathway activation [15] . We therefore evaluated IIS activity in body tissues of adult dilp mutants by measuring transcript levels of the translational regulator 4E-BP ( encoded by Thor ) , a direct target of dFOXO , which is induced when IIS is repressed and dFOXO is activated ( Figure 3A and 3B ) . As expected , 4E-BP transcript levels were up-regulated in dilp2–3 , 5 mutants ( Figure 3B ) , consistent with peripheral activation of dFOXO and reduced IIS . However , we did not observe significant up-regulation of 4E-BP transcript levels in dilp single mutants , except for dilp3 mutants , which showed slight up-regulation of 4E-BP levels in bodies but not in heads ( Figure 3A ) . Thus , the knock out of most individual DILPs did not result in systemic down-regulation of IIS that was detectable by measuring 4E-BP transcript levels . A possible explanation for this finding could be that DILPs act redundantly as part of a negative feedback system in which knock-out of one DILP is compensated by the up-regulation of others . To address the possibility of compensatory regulation we measured dilp transcript levels in dilp mutant flies ( Figure 3B and 3C ) . We did not detect any up-regulation of dilp2 or dilp3 transcript levels in dilp1 , 4 , 6 or 7 mutants , however , dilp5 was up-regulated in dilp2 and dilp2–3 mutants and dilp3 was up-regulated in dilp2 and dilp5 mutants ( Figure 3C ) , demonstrating compensatory transcriptional regulation among MNC-expressed DILPs . Intriguingly , dilp2 and dilp5 expression was down-regulated in dilp3 mutants ( Figure 3C and 3D ) , suggesting synergy in expression , with dilp3 acting as a positive regulator of dilp2 and dilp5 expression . Remarkably , while expression of dilp4 and dilp7 was not significantly changed in dilp2–3 , 5 mutants , the fat-body-expressed dilp6 gene was strongly up-regulated ( Figure 3B ) , suggesting the existence of a negative feedback system that acts to coordinate DILP expression between the MNCs in the central nervous system and peripheral tissues like the fat body ( Figure 3E ) . Interestingly , expression of the DILP binding protein ImpL2 , a negative regulator of IIS [32] , was down-regulated in dilp2–3 , 5 mutants ( Figure 3B ) , demonstrating that the negative feedback system is not restricted to the regulation of dilp transcription . The compensatory regulation among dilp genes may indicate that they act at least in part redundantly . However , each dilp gene is expressed in a tissue- and stage specific manner , suggesting that there is also diversification of their functions . In order to examine the in vivo function of individual dilp genes and to address whether and to what extent they act redundantly and synergistically to execute these functions , we initiated a systematic analysis of the dilp mutants , addressing phenotypes associated with MNC-ablation and reduced IIS , including egg-to-adult survival , development time , organismal growth , stress resistance , energy storage , lifespan and fecundity ( summarized in Table 1 ) . All seven dilp single mutants as well as dilp2–3 and dilp1–4 mutants were homozygous viable ( Table 1 ) . In contrast , dilp2–3 , 5 mutants showed sex-specific lethality; whereas homozygous mutant females showed normal viability , only 50–60% of dilp2–3 , 5 homozygous males developed into adult flies ( Table 1 ) . Survival was not further negatively affected in dilp7;2–3 , 5 mutant flies , but was reduced in dilp1–4 , 5 mutants . Intriguingly , animals that lacked all DILPs except for DILP6 ( dilp7;1–4 , 5 mutants ) still developed into adult flies . In contrast , combined knock-out of DILPs 2 , 3 , 5 and 6 caused complete lethality in males and females . This result indicates that DILP6 acts redundantly to MNC-expressed DILPs , consistent with the compensatory up-regulation of DILP6 transcript in dilp2–3 , 5 mutants ( Figure 2B ) . Lethality in combination with dilp2–3 , 5 mutants was observed for both dilp6 alleles , further evidence that dilp641 is a dilp6 loss-of function allele . dilp2 and dilp6 were the only single mutants that showed a delay in egg-to-adult development ( Figure 4A , Table 1 ) . Development time was only slightly further delayed in dilp1–4 mutants compared to dilp2 single mutants . In contrast , dilp2–3 , 5 mutants had a severe developmental delay ( Figure 4A ) , comparable to flies with ablated MNCs or DInR mutants [10] . The developmental delay was caused by delays in larval or pupal development , because dilp2–3 , 5 homozygous mutant embryos developed into first instar larvae at the same rate as wild type controls ( data not shown ) . dilp2–3 , 5 mutants also eclosed over a much longer period; all control flies eclosed within a day while dilp2–3 , 5 mutant flies continued to hatch over a period of almost ten days ( Figure 4A ) . dilp1–4 , 5 mutants , dilp7;2–3 , 5 mutants and dilp7 , 1–4 , 5 mutants had similar development times to dilp2–3 , 5 mutants ( Table 1 ) , suggesting that DILPs1 , 4 and 7 are not involved in the regulation of developmental timing . Organismal growth was analysed by measuring the body weight of adult flies ( Figure 4B and 4C , Table 1 ) . dilp 3 , 4 , 5 and 7 single mutants showed normal body weight . dilp1 and dilp2 mutants showed a slight reduction in weight ( Figure 4B ) , consistent with the shorter adult body length seen upon RNAi-mediated knockdown of DILP1 and DILP2 [17] . Although dilp1 mutants weighed less , they developed at the same rate as control flies , demonstrating that growth defects are not necessarily coupled with a delay in development . Intriguingly , dilp6 mutants showed the biggest reduction in body weight of all dilp single mutants ( Figure 4B ) . DILP6 resembles IGFs and is expressed at high levels in the fat body but not in the MNCs [6] , suggesting DILP6 to be an IGF-like peptide secreted by the fat body that promotes growth during larval-pupal development . dilp2–3 mutants weighed as much as dilp2 mutants and only a minor additional decrease in body weight was seen in dilp1–4 mutants ( Figure 4B ) , likely to be the result of the combined lack of DILP1 and DILP2 . In contrast , body weight of dilp2–3 , 5 mutants was severely reduced ( Figure 4C ) , and an even further reduction was observed in dilp1–4 , 5 mutants , which were approximately 50% smaller than controls ( Figure 4C ) . Lack of DILP7 did not result in a further decrease in body weight of either dilp2–3 , 5 or dilp1–4 , 5 mutants ( Figure 4C ) , suggesting that DILP7 does not contribute to the regulation of organismal growth . Oxidative stress and starvation resistance of dilp mutants was analysed by monitoring the survival of females on 20 mM paraquat and 1% agar , respectively . None of the dilp single mutants or the dilp2–3 mutants was more resistant to paraquat or starvation treatment ( Table 1 , Figure S4 , Figure S5A ) . Additionally , dilp single mutants did not show increased glycogen or lipid storage , with the exception of dilp6 mutants , which had slightly increased lipid levels ( Figure S5B , S5E ) . Whole body trehalose levels were increased in dilp2 mutants , but not changed in dilp3 or dilp5 mutants ( Figure S5D ) , consistent with the previous suggestion that stored trehalose levels are specifically regulated by DILP2 [18] . dilp1–4 mutants were slightly more resistant to paraquat and starvation treatment than controls , and dilp2–3 , 5 mutants were highly resistant to oxidative stress , as demonstrated by their increased survival under both paraquat and hydrogen peroxide treatment ( Figure S4 ) . However , and in contrast to MNC ablated flies , they were not resistant to starvation ( Figure S5 ) , even though they stored more energy in the form of glycogen ( Figure S5C ) and lipids ( Figure S5F ) . This finding suggests either that lack of DILPs other than DILP2 , 3 or 5 is causal for the increased starvation resistance of MNC-ablated flies , or that MNCs mediate starvation resistance independent of DILP function . The latter is consistent with the proposed function of the dARC1 protein in MNCs , which has been suggested to control the behavioural response to starvation , the lack of which might induce starvation resistance [33] . MNC-ablation experiments have suggested a role for DILPs in the determination of lifespan [11] , [34] . In particular DILP2 has been proposed by a number of studies to play an important role , because of its transcriptional down-regulation in mutant , long-lived flies [14]–[17] . However , this view has been challenged recently by the finding that RNAi-mediated knock-down of DILP2 is not sufficient to extend lifespan in flies [18] . We measured the lifespans of all seven dilp null mutants using female flies kept on standard food . We did not observe lifespan-extension in dilp1 , 3 , 4 , 5 , 6 or 7 mutants ( Figure S6A ) . However , in contrast to dilp2 RNAi hypomorphs , dilp2 null mutants were significantly longer-lived than controls ( Figure 5 ) . An increase between 8% and 13% in median lifespan was observed in four independent trials , two genetic backgrounds and in dilp2 mutants originating from independent homologous recombination events tested individually or as transheterozygotes ( Figure 5 and data not shown ) . Furthermore , a 9% extension of median lifespan was also observed for dilp2 mutant males , demonstrating that DILP2 is limiting for lifespan in both sexes . dilp2–3 mutants were also long-lived , although no more so than dilp2 mutants ( Figure 5 ) . Interestingly , while heterozygous dilp2–3 , 5 mutants were slightly long-lived , neither homozygous dilp2–3 , 5 mutants nor dilp1–4 mutants showed an increased median lifespan under standard food conditions ( Figure 5 , Figure S6A ) . However , maximum lifespan of homozygous dilp2–3 , 5 mutants was increased by 14% , as reported for the demographic aging of flies in which MNC were ablated early during development [35] . These findings might suggest that the strong reduction in insulin signalling in these mutants produced a general decrease in adult viability as well as a slowing down of the increase in death rates with age . The intracellular symbiont Wolbachia pipientis , a maternally transmitted bacterium , has been shown to modulate longevity in wild type and mutant stocks of Drosophila [36] , [37] and has recently been suggested to reduce the severity of IIS mutants by increasing IIS downstream of the DInR [38] . Therefore , we decided to test the influence of Wolbachia on lifespan and other fitness-related traits of dilp2–3 , 5 mutants ( Figure 6A–6E ) . Intriguingly , Wolbachia-positive wDah;dilp2–3 , 5 mutants were extremely long-lived , showing an increase on standard food in both median and maximum lifespan of 29% and 22% , respectively , compared to wDah controls ( Figure 6A ) . Lifespan extension was even more pronounced on a high yeast diet with an increase of up to 55% and 27% in medium and maximum lifespan , respectively ( Figure 7G ) . In contrast , Wolbachia had no effect on the lifespan of wild type flies ( Figure 6A ) , confirming that lifespan extension of dilp2–3 , 5 mutants is the result of a specific interaction between Wolbachia and the IIS pathway . However , Wolbachia did not affect IIS pathway activity as measured by 4E-BP expression . Although 4E-BP was up-regulated in dilp2–3 , 5 mutants ( Figure 3B ) , there was no significant difference in 4E-BP expression between dilp2–3 , 5 mutants or wild type flies , respectively , with and without Wolbachia ( Figure 6G ) . Some other consequence of insulin signaling must therefore mediate the effects of Wolbachia . Intriguingly , although Wolbachia is well known to manipulate the reproductive system of its host , we did not detect significant differences in fecundity between mutants or controls with or without bacterial infection ( Figure 6B ) , and nor did it have an effect on development time ( data not shown ) or energy storage ( Figure S5C , S5F ) . Wolbachia did affect growth , but not by increasing IIS; flies without Wolbachia infection had significantly lower body weights than flies carrying the bacterium ( Figure 4C ) , but this effect was seen in wild type and dilp2–3 , 5 mutants . Wolbachia infection status did not change survival of dilp2–3 , 5 mutants under starvation or hydrogen peroxide treatment ( Figure 6C and 6D ) , suggesting that oxidative stress resistance of wDah;dilp2–3 , 5 mutants is not causal for their increased lifespan . In contrast , Wolbachia-positive dilp2–3 , 5 mutants were more resistant to DDT treatment than Wolbachia-free dilp2–3 , 5 mutants ( Figure 6E ) , which suggests that xenobiotic resistance may contribute to their increased longevity . IIS has been implicated in the maintenance of germ-line stem cells ( GSC ) in Drosophila and over-expression of DILP2 suppressed GSC loss in adult females [13] . Consistently , dilp2 mutant females exhibited a significantly reduced lifetime egg-production ( −25% ) compared to control flies ( Figure S6B ) . However , fecundity was already reduced in young 3 day old dilp2 mutant females ( data not shown ) , suggesting that at least part of the phenotype is caused by developmental defects , consistent with results from flies with ablated MNCs [34] . A small decrease in fecundity was also observed in dilp3 ( −22% ) , dilp5 ( −18% ) , dilp2–3 ( −27% ) and dilp1–4 ( −14% ) mutants ( Figure S6B ) . In contrast , egg-production of dilp2–3 , 5 mutants was more severely reduced ( −69% , Figure 6B ) , suggesting that DILP2 , 3 and 5 can act redundantly in the control of egg-production . Notably , dilp2–3 , 5 females are not completely sterile like other strong IIS pathway mutants e . g . homozygous chico females [39] . Thus , the finding that dilp6 mutants exhibited the strongest reduction in fecundity of all dilp single mutant females ( −46% , Figure S6B ) suggests that egg-production is under the combined control of DILPs expressed in MNCs and the fat body . In contrast , fecundity was not significantly reduced in dilp1 , 4 and 7 mutants , the latter contrary to the suggestion that DILP7 may be a Drosophila relaxin [7] . DR and IIS have been suggested to act via overlapping mechanisms to extend lifespan in flies [19] , [22] , and the abundance of dilp5 mRNA is reduced in DR flies [20] , suggesting a possible role for DILP5 in the flies' responses to DR . To determine whether DILPs contribute to the DR response , we measured the lifespan of dilp2 , 3 and 5 mutant females ( Figure 7A–7C ) under DR . The dilp single mutants showed a normal DR response , with a peak in lifespan on 1 . 0x food and the shortest lifespan on 2 . 0x food observed for both mutants and controls . Except for starvation conditions ( 0 . 1x ) , dilp2 mutants were long-lived on all food types , suggesting that the lack of DILP2 causes lifespan-extension independent of yeast concentration ( Figure 7A ) . dilp5 mutants showed a normal DR response , demonstrating that DILP5 is not essential for DR mediated lifespan extension . However , lack of DILP5 may be compensated for by up-regulation of other DILPs on the higher yeast concentrations . We therefore measured DILP transcript levels in the dilp mutants on 2 . 0x food ( Figure 7D ) . Whereas the transcriptional regulation of DILP transcripts in dilp2 and dilp3 mutants was similar on low ( 1 . 0x ) and high ( 2 . 0x ) yeast concentrations ( compare Figure 7D to Figure 3C ) , in dilp5 mutants expression of DILP2 was up-regulated on 2 . 0x food ( Figure 7D ) , suggesting that the lack of diet-dependent DILP5 expression was compensated for by up-regulation of DILP2 . Thus , DILPs in the MNC can act redundantly to mediate the organismal response to DR . To test for redundancy , we measured the DR response of dilp2–3 , 5 mutant females in two independent trials using flies with ( Figure 7G and 7H ) and without Wolbachia ( Figure 7E and 7F ) . Wolbachia-free dilp2–3 , 5 mutants failed to show a normal response to DR . Instead , similar to chico1 mutants , their response was right shifted [22] , with the flies shorter-lived compared to controls on low but longer-lived on high yeast concentrations ( Figure 7E and 7F ) . The maximum lifespan of dilp2–3 , 5 mutants did not exceed the maximum that was achieved by DR alone , consistent with reduced IIS and DR extending lifespan through the same mechanisms . In addition , whereas wild-type flies showed a strong increase in egg-production between 1x and 2x food ( 62–81% ) , Wolbachia-free dilp2–3 , 5 mutants only laid 26% more eggs on the higher yeast concentration ( Figure 7E and 7F ) . Absence of these three DILPs therefore strongly attenuated the response of fecundity to DR . As in wild type flies only DILP5 expression was nutritionally regulated , these results suggest that the normal response to DR is mainly mediated by DILP5 . Interestingly , infection with Wolbachia modified the DR response of dilp2–3 , 5 mutants . In contrast to Wolbachia-free dilp2–3 , 5 mutants , these mutants showed extended lifespan at all food concentrations except for starvation , and their maximum lifespan far exceeded that achieved by DR treatment alone ( Figure 7G and 7H ) . In addition , the DR response of the Wolbachia-positive dilp2–3 , 5 mutant flies was greatly attenuated but not right shifted . Whereas control flies exhibited a DR-induced lifespan extension of 14–20% and an increase in egg production of 108–123% between 1x and 2x food , dilp2–3 , 5 mutants showed lifespan extension of only 2–6% and an increased egg production of 7–48% . Wolbachia infection status had no effect on the lifespan response to DR of wild type control flies , consistent with previous findings [40] . In conclusion , the DR response of Wolbachia-containing dilp2–3 , 5 mutants revealed both that these ligands mediate the responses to DR and that reduced IIS extends lifespan through mechanisms that both overlap with those of DR and through additional mechanisms that are independent of those at work in DR .
Wolbachia pipientis are maternally-inherited , obligate intracellular bacteria that are extremely widespread among wild and laboratory Drosophila populations [45] and their presence has been associated with parasitic and/or endosymbiontic modification of host fitness-related traits including lifespan [36] , [37] . Interestingly , specific interaction between Wolbachia strain and host genotype have been demonstrated , e . g . Wolbachia can suppress the sterility phenotype of sex-lethal mutants or modify the longevity of a long-lived Drosophila strain [37] , [46] . Here we show that longevity of dilp2–3 , 5 mutant flies is dependent on the presence of a maternally derived factor that can be removed by treatment with Tetracycline , most likely Wolbachia . The Tetracycline treatment itself is unlikely to have caused any negative effects because wild type flies with and without Wolbachia were phenotypically indistinguishable , except for their different body weight . Furthermore , dilp2–3 , 5 mutants had the same median lifespan as control flies in two Wolbachia-free genetic backgrounds , which demonstrates that the lack of lifespan extension is not specific to the outbred wDahT strain . Wolbachia-positive dilp2–3 , 5 mutants were extremely long-lived and had a prolonged survival under DDT treatment . Increased resistance to xenobiotic compounds has previously been associated with increased longevity [47] . For example , long-lived Little mice , mutant for the GH-releasing hormone receptor gene have been shown to be resistant to xenobiotic toxicity and show concerted up-regulation of xenobiotic detoxification genes [48] , [49] . Furthermore , long-lived IIS mutant C . elegans and Drosophila also show increased expression of genes involved in xenobiotic metabolism [50] , suggesting that xenobiotic resistance may contributes to the lifespan extension of wDah , dilp2–3 , 5 mutant flies . However , compared to control flies Wolbachia-free dilp2–3 , 5 mutants also showed increased survival under DDT treatment ( Figure 6E ) , demonstrating that increased xenobiotic resistance alone was not sufficient to increase lifespan , suggesting that in addition other mechanisms contribute to the Wolbachia-dependent lifespan extension of dilp2–3 , 5 mutants . However , the molecular mechanisms by which Wolbachia influences its host are currently unknown . Notably , lifespan extension of dilp2 and dilp2–3 double mutants was not dependent on the presence of Wolbachia ( Figure 5 ) , demonstrating that Wolbachia is not in general essential for lifespan extension due to reduced IIS in Drosophila . However , this finding raises the question why Wolbachia is essential for lifespan extension in one IIS mutant but not in another . There may exist an optimal range of down-regulation of IIS pathway activity in order to extend lifespan . In contrast to the relative mild phenotypes of dilp2 or dilp2–3 mutants , the combined loss of DILP2 , 3 and 5 causes severe developmental and metabolic phenotypes , which may be detrimental to the flies , and Wolbachia may attenuate the expressivity of the dilp2–3 , 5 mutant phenotype . Recently , it has been suggested that Wolbachia acts to increase insulin signaling downstream of the DInR and thereby attenuates the phenotype of flies overexpressing a dominant negative DInR [38] . In contrast to these results we did not observe changes in IIS pathway activity when comparing expression of the IIS target 4E-BP in flies with or without Wolbachia . In addition , we found no difference in egg-to-adult survival , development time , energy storage , stress resistance or fecundity between dilp2–3 , 5 mutants with or without Wolbachia . This observation suggests either that lifespan , in contrast to the other traits , is very sensitive to even small changes in IIS activity or that Wolbachia mediates lifespan extension of dilp2–3 , 5 mutants by another mechanism . The interaction between Wolbachia and its Drosophila host are complex and dependent both on the Wolbachia strain as well as the genetic background of the fly line . In addition , rapid co-evolution between Wolbachia and its host has been demonstrated [51] . In our study we analysed the interaction between IIS and one Wolbachia strain in the context of its natural host , the outbred wDahomey line . For future studies it will be interesting to determine whether this interaction is specific for Dahomey flies and its co-evolved Wolbachia strain , or whether other Wolbachia strains and/or other Drosophila wild type lines have the same effect on IIS . Dietary restriction , the reduced availability of nutrients without malnutrition , extends lifespan in a wide variety of organisms including worms , flies and mammals . However , the underlying molecular pathways mediating the effect of DR on lifespan are still elusive . In Drosophila , IIS and DR have been suggested to act through overlapping mechanisms , based on the DR response of chico mutants , which are short-lived on low food concentrations but long-lived on high food concentrations [22] . However recently the relevance of IIS for the response to DR in Drosophila has been challenged by the finding that flies mutant for the downstream target of IIS the transcription factor dFOXO are short-lived , but respond equally well to DR as control flies [19] , [20] . Here we present evidence that DR in Drosophila is mediated by the up-stream ligands of IIS , DILPs , expressed in the MNCs . Although dilp single mutants responded as strongly to DR as did control flies , the DR response of dilp2–3 , 5 mutants was either severely attenuated or completely blocked , depending on the presence of Wolbachia , suggesting that DILPs can act redundantly in mediating the response to DR . Additionally , transcript levels of DILP5 were found to be regulated in a diet-dependent manner . Whereas DILP2 and DILP3 transcript levels remained constant across diets , the abundance of DILP5 mRNA was reduced in dietary restricted flies [20] , suggesting that , under normal conditions , the response to DR is mediated by DILP5 . However , when DILP5 is missing other DILPs may compensate the lack of diet-induced DILP5 expression , consistent with the up-regulation of DILP2 transcript in dilp5 mutants on high yeast food only . RNAi-mediated knock-down of DILP3 has been shown to reduce DILP5 transcript levels and to block diet-dependent changes in DILP5 transcription and these flies respond normally to DR treatment [20] , which we could confirm using the dilp3 null mutant flies . However , this finding does not exclude a function for DILP5 in the response to DR , as DILP5 peptide is still present and could be regulated on the level of protein stability or secretion . Thus , ligands of IIS mediate the changes in longevity seen under DR conditions in Drosophila , which raises the question why do mutants of the IIS downstream effector dFOXO show a normal response to DR . One explanation could be that dFOXO is involved in the response to DR under normal conditions but in its absence another pathway mediates the life span extension seen upon DR treatment . Fat body specific over-expression of dFOXO extends lifespan in a nutrient dependent manner , which would be consistent with a role of dFOXO in DR [19] . In C . elegans , the forkhead transcription factor pha-4 , a Foxa orthologue , was shown to be required for lifespan extension under DR [52] and its fly orthologue would therefore be a candidate to mediate DR in the fly , redundant to IIS . The Target of Rapamycin ( TOR ) pathway has been linked to the determination of lifespan in flies and worms and lifespan-extension by decreased TOR signalling is dependent on nutritional conditions , suggesting a possible link between the TOR pathway and DR [53] , [54] . The IIS pathway regulates TOR activity through the protein kinase AKT ( PKB ) , an IIS component downstream of DILPs and Chico but upstream of dFOXO [1] . Thus , in Drosophila upstream IIS components such as DILPs and Chico may mediate the response to DR via the TOR pathway but not through the IIS downstream effector dFOXO . Strong evolutionary conservation of dilp gene family membership and sequence differentiation has thus been accompanied by functional differentiation , redundancy and synergy between DILPs , and these features may themselves be of evolutionary advantage .
dilp mutants were generated by ends-out homologous recombination according to the methods described in [30] , [55] . All fly stocks are summarized in Table S1 . Donor constructs used for targeting dilp genes are summarized in Table S2 . DNA fragments homologous to approximately 4 kb of dilp gene flanking sequences were amplified by PCR and subsequently cloned into the pW25 vector [55] . pw25 was obtained from the Drosophila Genomics Resource Center , ( Bloomington , Indiana , USA ) . Long-range PCR was done using Takara LA Taq polymerase ( Lonza , UK ) on BAC clone DNA as template . BAC clones RP98-7A5 for dilp1–5 and RP98-32E2 for dilp7 were obtained from the BACPAC Resource Center ( Oakland , California , USA ) . Primer sequences and restriction sites used for subcloning into pw25 are summarized in Table S3 . DILP donor constructs were full-length sequenced and checked for base pair substitutions in gene coding sequences before used to generate transgenic flies . DNA constructs were transformed into the germline of Drosophila melanogaster by P-element-mediated germ line transformation using the Best Gene Drosophila Embryo Injection Services , ( Chino Hills , California , USA ) . Crosses for ends-out homologous recombination were carried out according to the rapid targeting scheme [56] . Subsequently , the whitehs marker gene was genetically mapped and homologous recombination events were identified by PCR . In order to generate dilp2–3 , 5 and dilp1–4 , 5 mutants , homologous recombination was done using flies carrying the dilp5 donor construct and hs-FLP , hs-SceI in the dilp2–31 and dilp1–41 mutant background , respectively . y1 P{y[+mDint2] w[BR . E . BR] = SUPor-P}KG04972 [57] flies that carry an KG-element transposon-construct integration in the 5′upstream region of the dilp6 gene at chromosome X position 2 , 229 , 002 , corresponding to position -2831 relative to the putative start ATG in dilp6 exon 2 , were obtained from the Bloomington Drosophila stock center , Indiana , USA . dilp6 deletion mutants were generated by conventional P-element imprecise excision resulting in the small dilp641 deletion and the large dilp668 deletion . Both dilp6 alleles contain residual P-element sequence , which impeded the exact mapping of the 3′ breakpoints at the sequence level . By PCR analysis the 3′ breakpoint in dilp641 could be mapped to the first dilp6 intron , i . e . dilp641 mutant flies lack the first dilp6 exon but contain the full dilp6 ORF . dilp668 mutants carry a deletion that covers at least 14 kb of genomic sequence including 5 annotated genes: dilp6 , CG33218 , CG2854 , CG14050 , CG34052 . dilp mutants were backcrossed for at least ten generations into two different wild-type stocks , the inbred lab strain w1118 and the outbred strain white DahomeyT ( wDahT ) . dilp6 mutants were only backcrossed into the wDahT background . Naturally , Dahomey flies carry the intracellular bacterium Wolbachia . wDahT flies were generated by treating white Dahomey ( wDah ) flies with Tetracycline to remove Wolbachia . The presence of Wolbachia was checked by PCR using a Wolbachia specific primer combination ( wsp-81F/wsp-691R , Table S3 and Figure 6F ) . w1118 flies do not contain Wolbachia and were therefore not treated with Tetracycline . Unless stated otherwise all experiments were done using Wolbachia-free flies . Flies carrying Wolbachia were generated by backcrossing dilp mutants into the original wDah background . All backcrossed stocks were maintained in large numbers in culture bottles on 1 , 0 SY-A medium at 25°C on a 12L∶12D cycle . Experimental flies were generated by crossing age-matched heterozygous mutants and wild type , heterozygous and homozygous mutant progeny were collected from the same bottle based on eye colour . ( Note: as homozygous dilp5 mutant males could not unambiguously be identified by their eye colour they were not used for experiments . ) Eggs were collected during a six-hour time window and the same volume of embryos was transferred to each rearing bottles ( 1 , 0 SY-A ) ensuring standard larval density . Flies eclosed during a 12 h time window were transferred to new bottles ( 1 , 0 SY-A ) and left for 48 h to mate ( once mated ) . Subsequently , flies were sorted under brief CO2 anaesthesia and transferred to vials . All experiments were conducted at 25°C on a 12 h∶12 h light:dark cycle at constant humidity ( 65% ) . For lifespan experiments , flies were maintained in vials at a density of 10 flies per vial on 1 . 0x SYA medium . Flies were transferred to new vials every two to three days and the number of dead flies was counted . DR experiments were done according to the optimised DR protocol described in [58] . Fly media used for DR experiments are summarized in Table S4 . For fecundity assays mated females were kept at a density of 5 females per vial . Eggs were collected during 16–20 hour periods during the first 3–4 weeks . The number of eggs laid per vial at each time point was counted . Data are reported as the cumulative number of eggs laid per average female ± SEM over the whole period . For DR experiments egg numbers were collected from the lifespan vials . Development time was analyzed by crossing heterozygous mutant flies and collecting eggs over a period of 3 h . Embryos were allowed to hatch and first instar larvae were hand picked and transferred 50 per vial on standard food . When adult flies started to hatch the number of eclosed homozygous , heterozygous and w control flies was counted in regular intervals . For body weight determination flies were briefly anaesthetized on ice and weighted individually or in pairs on a ME235S analysis balance ( Sartorius Mechantronics ) . For stress tests 20 once mated females per vial were kept on 1x SY-A food for 8–10 days before the assay and transferred every other day . Media used for oxidative stress , starvation and DDT assays are summarized in Table S4 . For DDT treatment time to knock-down was measured . Organismal lipid , glycogen and trehalose content of 8–10 day old females was quantified as described [18] , [59] . DILP2 Western blots were done as described [18] . Immunohistochemistry on fly brains of 8–10 day old females was done according to the procedures described in [11] . The following primary antibodies were used: anti-DILP2 ( rabbit ) , anti-DILP3 ( rabbit ) , anti-DILP4 ( rabbit ) and anti-DILP5 ( rat ) . Pictures were taken using an LSM 510 confocal microscope ( Zeiss ) . Total RNA was extracted from 25 fly heads or 25 bodies using Trizol ( Gibco ) according to the protocol of the manufacturer . cDNA was prepared using the Superscript II RT system ( Invitrogen ) and oligo ( dT ) primer . Quantitative RT-PCR was performed on a PRISM 7000 sequence detection system using SYBR green master mix ( Applied Biosystems ) and four independent RNA extractions per genotype . Results were calculated according to the standard curve method and normalized using act primers . Primers used for Q-RT-PCR are summarized in Table S3 . Lifespan and stress assays were recorded using Excel and were subjected to survival analysis ( log rank test ) and presented as survival curves . Fecundity data were analysed using the Wilcoxon rank test . Q-RT-PCR , glycogen , trehalose and lipid data were recorded in Excel and analysed using Student's t-test .
|
The insulin/IGF signalling ( IIS ) pathway plays key roles in growth , metabolism , reproduction , and longevity in animals as diverse as flies and mammals . Most multicellular animals contain multiple IIS ligands , including 7 in the fruit fly Drosophila melanogaster ( DILP1-7 ) , implying that the diverse functions of IIS could in part be mediated by the functional diversification of the ligands . Although Drosophila is a prime model organism to study IIS , knowledge about the function of individual DILPs is still very limited due to the lack of gene-specific mutants . Therefore , we generated mutants for all 7 dilp genes and systematically analyzed their phenotypes . We show that loss of DILP2 extends lifespan and describe a novel role for DILP6 in growth control . Furthermore , we show that DILPs are evolutionary conserved and can act redundantly , supporting the hypothesis that functional redundancy itself can be of evolutionary advantage . We also describe a specific interaction between IIS and the endosymbiontic bacterium Wolbachia in lifespan regulation . This finding has implications for all longevity studies using IIS mutants in flies and offers the opportunity to study IIS as a mechanism involved in host/symbiont interactions . Finally , we show that DILPs mediate the response of lifespan and fecundity to dietary restriction .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology/aging",
"diabetes",
"and",
"endocrinology"
] |
2010
|
Molecular Evolution and Functional Characterization of Drosophila Insulin-Like Peptides
|
The lifespan and activity of proteins depend on protein quality control systems formed by chaperones and proteases that ensure correct protein folding and prevent the formation of toxic aggregates . We previously found that the Arabidopsis thaliana J-protein J20 delivers inactive ( misfolded ) forms of the plastidial enzyme deoxyxylulose 5-phosphate synthase ( DXS ) to the Hsp70 chaperone for either proper folding or degradation . Here we show that the fate of Hsp70-bound DXS depends on pathways involving specific Hsp100 chaperones . Analysis of individual mutants for the four Hsp100 chaperones present in Arabidopsis chloroplasts showed increased levels of DXS proteins ( but not transcripts ) only in those defective in ClpC1 or ClpB3 . However , the accumulated enzyme was active in the clpc1 mutant but inactive in clpb3 plants . Genetic evidence indicated that ClpC chaperones might be required for the unfolding of J20-delivered DXS protein coupled to degradation by the Clp protease . By contrast , biochemical and genetic approaches confirmed that Hsp70 and ClpB3 chaperones interact to collaborate in the refolding and activation of DXS . We conclude that specific J-proteins and Hsp100 chaperones act together with Hsp70 to recognize and deliver DXS to either reactivation ( via ClpB3 ) or removal ( via ClpC1 ) depending on the physiological status of the plastid .
Organelles like mitochondria and plastids play fundamental roles in all eukaryotic organisms . In particular , plastids were acquired by a symbiosis between photosynthetic cyanobacteria and eukaryotic cells . Today , plastids ( like mitochondria ) are intimately integrated into the metabolism of plant cells but they still remain as separate functional entities that regulate their own biochemistry by relatively independent mechanisms . An important part of this regulation relies on the effective control of plastidial enzyme activities . Most of the enzymes required for plastidial metabolism are encoded by nuclear genes , synthesized in precursor form in the cytosol , and transported into plastids using energy-dependent import machineries [1] . Following import , specific proteases cleave the transit peptides and complex networks of plastidial chaperones ensure proper folding , assembly , or suborganellar targeting of the mature proteins . Chaperones and proteases are also essential components of the protein quality control ( PQC ) system that promotes the stabilization , refolding , or degradation of mature proteins that lose their native conformation and activity after metabolic perturbations or environmental challenges such as excess light , temperature peaks , oxidative stress or nutrient starvation [2 , 3] . While plant plastids contain many groups of prokaryotic-like chaperones ( such as Hsp70 and Hsp100 ) and proteases ( including Clp , Lon , Deg , and FstH ) , their specific targets and PQC-related roles remain little studied [1–4] . Due to the presence of plastids , plants have biochemical pathways that are not found in other eukaryotic kingdoms . For example , isoprenoid precursors are produced by the methylerythritol 4-phosphate ( MEP ) pathway in bacteria and plant plastids , whereas animals and fungi synthesize these essential metabolites using a completely unrelated pathway which is also used by plants to produce cytosolic and mitochondrial isoprenoids [5 , 6] . MEP-derived isoprenoids include compounds essential for photosynthesis ( such as carotenoids and the side chain of chlorophylls , tocopherols , plastoquinone and phylloquinones ) and growth regulation ( including the hormones gibberellins , cytokinins , strigolactones and abscisic acid ) . Many plastidial isoprenoids also have nutritional and economic relevance [6] . All MEP pathway enzymes are located in the plastid stroma [5 , 7] . While transcriptional regulation of genes encoding biosynthetic enzymes is known to exert a coarse control of the MEP pathway , fine-tuning of metabolic flux appears to rely on post-transcriptional or/and post-translational regulation of enzyme levels and activity [8–12] . This is most evident for deoxyxylulose 5-phosphate synthase ( DXS ) , the homodimeric enzyme that catalyzes the first step of the pathway . Metabolic control analysis calculations confirmed that DXS is the enzyme with the highest flux control coefficient ( i . e . the main rate-determining step ) of the MEP pathway [13] . Consistent with this prime regulatory role , DXS activity is tightly regulated by several post-translational mechanisms [10–12] . In particular , DXS enzymatic activity is allosterically inhibited by MEP pathway products [14 , 15] , which also repress DXS protein accumulation [8 , 14 , 16–18] . Mathematical modeling recently showed that the post-translational control of DXS protein abundance and enzyme activity is crucial for the adjustment of the MEP pathway flux to persistent changes in environmental conditions , such as substrate supply or product demand [18] . Despite the central relevance of this type of regulation , little is known about the molecular mechanisms behind it . We previously showed that the Arabidopsis thaliana J-protein J20 interacts with inactive forms of DXS to deliver them to the Hsp70 chaperone for eventual activation ( which involves folding or refolding ) or degradation ( which involves unfolding ) [19] . However , the particular protease involved and the specific components of the two J20-dependent antagonistic pathways remained unknown . Here we show that DXS is primarily degraded by the Clp protease complex through a pathway involving J20 and Hsp100 chaperones of the ClpC type . We also demonstrate that Hsp70 can physically interact with ClpB3 , another plastidial Hsp100 chaperone , to promote the activation of non-functional DXS enzymes .
The main protease families involved in the degradation of terminally damaged or surplus proteins in plastids are Clp , Lon , Deg , and FtsH , all of them of prokaryotic origin [3 , 4] . We and others have previously shown that Arabidopsis mutants with a decreased activity of the stromal Clp protease complex display an accumulation of several MEP pathway enzymes , including DXS [20–24] . However , whether other plastidial proteases involved in PQC networks could also contribute to DXS degradation in the stroma remains unexplored . Several functional Lon homologues are found in Arabidopsis , but only Lon1 [25] and Lon4 [26] have been localized to chloroplasts , where they are attached to the stromal side of thylakoids . The Deg gene family in Arabidopsis contains 16 members , with 5 of them experimentally confirmed to be localized in chloroplasts [27] . From these , the isoforms Deg1 , Deg5 and Deg8 were found in the thylakoid lumen , whereas Deg2 and Deg7 were detected in the stromal side [28 , 29] . FtsH proteases are encoded by 12 genes in Arabidopsis , and 9 of them can be found in chloroplasts [30] . The four major chloroplast isomers ( FtsH2 , FtsH5 , FtsH8 and FtsH1 , in order of abundance ) have been shown to reside in the thylakoid membrane with their catalytic domain facing the stromal side [31–33] . DXS protein levels were examined by immunoblot analysis in Arabidopsis wild-type ( WT ) plants and single mutants defective in plastidial Lon , Deg , or FtsH isomers ( Fig 1 and S1 Table ) . As a control , we included the Clp protease mutant clpr1 , which displays a reduction of other subunits of the Clp proteolytic core [34] but increased DXS protein levels [20] . As shown in Fig 1A , DXS protein levels in the analyzed mutants were similar to those in WT plants with only three exceptions . Lines defective in Lon1 and Deg7 showed a decreased accumulation of the protein compared to the WT , whereas DXS levels were only increased in the clpr1 mutant ( Fig 1A ) . No changes in DXS transcript levels were detected in any of the mutant lines ( Fig 1A ) . Since enzyme levels would be expected to be post-translationally increased ( but not decreased ) in the mutants impaired in DXS-degrading proteases , we conclude that the Clp complex is likely the primary protease for DXS removal . The contribution of Lon , Deg , or FtsH proteases , however , cannot be fully discarded as we only tested mutants for individual isoforms of those other proteases and it is possible that different isoforms may have redundant functions . Clp proteases are found in almost all bacteria and endosymbiotic organelles ( mitochondria and plastids ) . In bacteria ( S1 Fig ) , they are formed by a barrel-like catalytic core of two heptameric rings of proteolytic subunits ( ClpP ) and a dynamically interacting hexameric ring of Hsp100 chaperones ( ClpA and ClpX in Escherichia coli; ClpC , ClpX , and ClpE in Bacillus subtilis ) that unfold substrates for translocation into the proteolytic chamber [35] . Additionally , interaction of Hsp100 members with adaptor proteins ( such as ClpS and SspB in E . coli and ClpS , MecA , McsB , and others in B . subtilis ) enhance or expand substrate specificity [35–37] . In plant plastids , the Clp protease is more complex [38 , 39] but the basic components are conserved ( S1 Fig ) . It presents a protease core ( formed by two heptameric rings of plastome encoded ClpP1 and nuclear-encoded ClpP3-P6 and ClpR1-R4 subunits ) stabilized by two plant-specific subunits ( ClpT1-T2 ) . The Arabidopsis homologues of the bacterial ClpA and ClpC unfolding chaperones are ClpC1 , ClpC2 , and ClpD . A ClpS adaptor is also found in chloroplasts [40] , where it might form a plant-specific binary adaptor complex with the ClpF protein [41] . The possibility of other pathways delivering proteins to the Clp protease , however , remains open . As shown above ( Fig 1A ) , DXS levels increase in mutants defective in Clp protease activity such as clpr1 [20] . If DXS is targeted to the Clp protease for degradation , we would also expect a post-translational upregulation of DXS enzyme levels in mutants impaired in the adaptors and chaperones that deliver the protein to the Clp catalytic core . A systematic analysis of such mutants ( clps , clpc1 , clpc2 , clpd , clpt1 , and clpt2 ) showed that only those defective in ClpC1 accumulated higher levels of DXS protein than WT plants ( Fig 1B and S2 Fig ) . Quantification of DXS-encoding transcripts in the same mutant lines showed WT levels in all cases ( Fig 1B ) , confirming that the observed accumulation of DXS polypeptides in ClpC1-defective lines was not a consequence of increased gene expression . It has been proposed that the two Arabidopsis ClpC paralogs ClpC1 and ClpC2 perform similar if not identical functions in the chloroplast [42] . However , proteolytic assays with known Clp protease substrates only showed a greatly reduced degradation rate in clpc1 plants [42] , which showed the strongest reduction in total ClpC content ( Fig 1B and S2 Fig ) . Estimation of DXS degradation rates upon treating WT and mutant plants with the protein synthesis inhibitor cycloheximide also showed a slower proteolytic removal of DXS polypeptides in clpc1 mutants ( Fig 2A ) . As expected , a defective Clp catalytic core in the clpr1 mutant led to similarly reduced DXS degradation rates ( Fig 2A ) , again supporting our conclusion that DXS is a target for this proteolytic complex . To confirm whether DXS might be a ClpC1 substrate , tagged versions of the Arabidopsis proteins ( DXS-GFP and ClpC1-MYC ) were overproduced in Nicotiana benthamiana leaves by agroinfiltration and co-immunoprecipitation assays were next performed . As shown in Fig 2B , these assays confirmed that DXS and ClpC1 can indeed interact . Together , we conclude that DXS might be mainly unfolded by ClpC1 for degradation by the Clp proteolytic core . Recent results have shown that client proteins of the stromal Clp protease are recognized and delivered to ClpC chaperones by ClpS and ClpF adaptors [40 , 41] . While DXS might actually be a target of ClpS in bacteria [43] , a wild-type phenotype in terms of DXS protein levels was observed in Arabidopsis plants defective in the proposed chloroplast adaptors ( Fig 1B ) [40 , 41] . Although ClpC could possibly directly deliver client proteins to the Clp protease without the need of an adaptor , we reasoned that further substrate specificity should be achieved using an alternative ClpS/ClpF-independent adaptor system . Our previous work showed that inactive forms of DXS are recognized by J20 , a J-protein adaptor that delivers them to the Hsp70 chaperone [19] . Computational analysis of the Arabidopsis DXS monomer with the Aggrescan3D algorithm revealed the presence of several aggregation-prone clusters ( S3 Fig ) . Consistent with the conclusion that DXS tends to aggregate and that J20 prevents its aggregation , GFP-tagged DXS proteins accumulate in plastidial speckles that are larger in j20 plants ( S4 Fig ) [19] . In addition , the endogenous DXS enzymes are less accessible to proteinase K cleavage in the j20 mutant ( S4 Fig ) , again suggesting that DXS aggregation is increased in the absence of J20 , likely because the delivery of aggregated ( and hence inactive ) DXS proteins to the Hsp70 chaperone is impaired . The main role of Hsp70 is actually to prevent the formation of toxic aggregates of damaged proteins and , together with Hsp100 chaperones , promote their solubilization [44–50] . However , Hsp70 chaperones also facilitate the transfer of irreparably damaged client proteins to proteolytic systems [49 , 51–53] . For example , cytosolic Hsp70 is involved in the degradation of Arabidopsis protein clients by the eukaryotic 26S proteasome [51] . Despite the absence of conserved domains for direct interactions between Hsp70 and ClpC-type Hsp100 proteins ( S5 Fig ) [36 , 45 , 46] , co-immunoprecipitation experiments showed that both chaperones can be found together in the chloroplast envelope [54 , 55] . It is therefore possible that Hsp70 and ClpC might interact either directly ( using unidentified chaperone binding motifs ) or indirectly ( via third partners ) to participate in PQC events at the stromal side of the inner envelope membrane [1 , 42 , 56 , 57] . Because in Arabidopsis the two plastidial isoforms of Hsp70 ( Hsp70 . 1 and Hsp70 . 2 ) and ClpC ( ClpC1 and ClpC2 ) are also found in the stroma [42 , 58] , we reasoned that Hsp70 and ClpC proteins might collaborate to deliver DXS to the Clp protease using J20 as an adaptor . Interestingly , overexpression of J20 in transgenic Arabidopsis plants leads to decreased DXS protein levels , whereas loss of J20 function causes a reduced degradation rate of the enzyme ( Fig 2C ) [19] . Since both the J20 adaptor and ClpC chaperones are involved in the control of DXS degradation , we next tested whether they might function in the same pathway . We followed a genetic strategy based on comparing the DXS accumulation phenotype of single mutants defective in either J20 or ClpC1 with that of double j20 clpc1 mutants ( Fig 3 ) . All three mutants accumulated higher levels of DXS proteins ( but not transcripts ) compared to WT plants . In particular , DXS levels increased ca . 2-fold in j20 plants and 4-fold in single clpc1 and double j20 clpc1 mutants ( Fig 3A ) . The absence of an additive or synergistic phenotype in the double mutant supports the conclusion that J20 and ClpC1 actually function in the same pathway delivering DXS to degradation in Arabidopsis plastids . Such a ClpS/ClpF-independent pathway could potentially be functioning for other plastidial clients of J-proteins . However , the lack of bona-fide substrates for other plastidial J-proteins prevents to experimentally testing this possibility at the moment . The results described above suggest that damaged DXS polypeptides might be recognized by J20 and then delivered to Hsp70 and ClpC chaperones for unfolding and degradation by the Clp proteolytic core . But unlike that observed for J20-defective mutants [19] , DXS accumulates in a mostly active form in clpc1 mutants ( Fig 3B ) . Thus , measurement of DXS activity in plant extracts showed increased total activity but unchanged specific activity ( i . e . relative to protein levels ) in clpc1 compared to WT controls ( Fig 3B ) . Estimation of DXS activity in planta by quantification of the resistance to clomazone ( CLM ) , a specific DXS inhibitor [19 , 59 , 60] further supported the presence of higher DXS activity levels ( i . e . increased resistance to the inhibitor ) in the clpc1 mutant , opposite to the increased sensitivity detected in the case of j20 plants ( Fig 3C and S6 Fig ) . To reconcile these results , we propose that loss of J20 causes an accumulation of aggregated ( i . e . enzymatically inactive ) DXS because the protein cannot be normally reactivated ( refolded ) or discarded ( degraded ) . By contrast , accumulation of active DXS enzyme when ClpC activity decreases might be due to the existence of a functional pathway to disaggregate and refold the excess protein that cannot be degraded in the clpc1 mutant . The participation of J20 in such a putative reactivation pathway is supported by the observation that loss of J20 function in degradation-impaired ( ClpC1-defective ) j20 clpc1 plants results in a higher proportion of inactive DXS enzyme ( Fig 3B ) and hence a reduced resistance to CLM ( Fig 3C and S6 Fig ) compared to the single clpc1 mutant . Work in different systems has shown that Hsp70 can be assisted by Hsp100 chaperones of the ClpB type in the solubilization of toxic aggregates of damaged proteins [45–50 , 61–64] . ClpB3 is the only ClpB-type chaperone found in Arabidopsis plastids [44] . Unlike the rest of plastidial Hsp100 chaperones present in this plant ( ClpC1 , ClpC2 , and ClpD ) , ClpB3 lacks the IGF motif ( or ClpP-loop ) required for interaction with proteolytic subunits of the Clp core but it harbors a domain responsible for the interaction with Hsp70 chaperones ( S5 Fig ) [36 , 45 , 46] . Interestingly , the levels of ClpB3 were increased in mutants defective in Clp protease subunits , including ClpC1 [21 , 22 , 24 , 40 , 65] ( Figs 1B and 3A and S2 Fig ) , suggesting that ClpB3 might contribute to mitigate protein folding stress caused by a defective Clp protease activity . In agreement , impairment of both ClpB3 and Clp protease activity results in a seedling lethal phenotype [22] . Based on these data , we speculated that ClpB3 might also participate in the DXS reactivation pathway mediated by J20 and Hsp70 chaperones . To evaluate this possibility , we first analyzed DXS protein levels and activity in ClpB3-defective Arabidopsis plants ( Fig 4 ) . If ClpB3 promotes DXS protein disaggregation ( and hence activation ) , it was expected that clpb3 mutants would show a transcription-independent accumulation of inactive forms of DXS , assuming that the degradation rate of J20-delivered proteins would remain constant . Indeed , clpb3 plants showed a WT rate of DXS degradation ( Fig 2C ) but an enhanced accumulation of DXS enzyme without changes in transcript levels ( Fig 4A ) . Also as predicted by our model , the specific activity of the DXS protein found in the ClpB3-defective mutant was much lower than that measured in WT plants ( Fig 4B ) . Loss of both ClpB3 and J20 activities in the double j20 clpb3 mutant resulted in an even higher accumulation ( Fig 4A ) of mostly inactive DXS protein ( Fig 4B ) , presumably because the absence of J20 prevents the targeting of non-functional enzymes to ClpC for eventual degradation by the Clp protease . The dramatic phenotype displayed by single clpb3 and double j20 clpb3 mutant plants ( Fig 4C ) [66] prevented the reliable quantification of their CLM resistance . In any case , the available data suggests that when the proteolytic degradation of inactive ( e . g . aggregated ) forms of DXS delivered to the Clp protease by J20 via ClpC is impaired ( e . g . in clpr1 and clpc1 mutants ) , an increase in ClpB3 levels promotes the disaggregation and activation of the enzyme , eventually resulting in higher levels of enzymatically active DXS . When J20 activity is missing , however , inactive DXS forms cannot be properly reactivated via ClpB3 ( as deduced from the similar levels of DXS protein but lower proportion of active enzyme found in the double j20 clpc1 mutant compared to the single clpc1 line; Fig 3 ) or degraded via ClpC ( as deduced from the increased levels of inactive DXS protein present in double j20 clpb3 plants compared to the single clpb3 mutant; Fig 4 ) . As described above , the mechanistic basis for the collaboration between J20 , Hsp70 , and ClpC chaperones is currently unknown . However , the presence of a Hsp70-binding motif in the amino acid sequence of ClpB3 ( S5 Fig ) suggests that plastidial Hsp70 isoforms might be able to directly interact with ClpB3 to synergistically activate damaged DXS proteins recognized by the J20 adaptor . In agreement with this possibility , the ClpB3 protein was efficiently immunoprecipitated from WT extracts using an anti-Hsp70 serum ( Fig 4D ) . When a similar experiment was performed with the Arabidopsis hsp70 . 2 mutant , previously shown to contain lower amounts of plastidial Hsp70 proteins than the WT [58] , the level of immunoprecipitated ClpB3 protein was concomitantly decreased ( Fig 4D ) . These results confirm that plastidial Hsp70 isoforms can be found together with ClpB3 in Arabidopsis chloroplasts , providing a mechanistic frame for the observed collaboration between these two families of chaperones in the J20-mediated activation of DXS . The results described above are consistent with a model involving the participation of ClpB3 and ClpC1 on opposite pathways resulting in either reactivation or degradation , respectively , of inactive DXS proteins recognized by the Hsp70 adaptor J20 . Under normal growth conditions , the levels of ClpB3 transcripts and protein are lower than those of ClpC1 ( S7 Fig ) [42 , 67] . However , ClpB3 transcript levels have been shown to strongly increase upon exposure to high temperatures [66 , 68 , 69] whereas virtually no changes in RNA or protein levels have been detected for ClpC1 or ClpC2 in response to heat or other types of stress , including cold , drought , salt , and oxidative stress [66 , 70] . The ratio between plastidial ClpB3 and ClpC1 chaperones ( and hence the potential capacity to reactivate damaged or/and aggregated DXS polypeptides ) could therefore increase when plants are challenged with at least some types of stress ( S7 Fig ) . DXS-derived isoprenoids such as carotenoids and tocopherols protect plants against oxidative stress , whereas others ( including chlorophylls and prenylated quinones ) are essential for photosynthesis . Therefore , a decreased production of these isoprenoids ( e . g . upon down-regulating DXS activity ) is expected to trigger a stress response . We observed that a specific reduction in DXS activity in Arabidopsis WT plants germinated and grown in the presence of CLM caused an increased accumulation of ClpB3 but not ClpC chaperones compared to controls grown in the absence of inhibitor ( Fig 5 ) . A similar ClpB3 protein accumulation response was also observed in mutants with a defective MEP pathway ( Fig 5 ) . As previously observed [8 , 14 , 16–18] , the pharmacological or genetic blockage of the pathway also resulted in increased accumulation of DXS protein . Most interestingly , the DXS and ClpB3 accumulation response was detected as soon as 5 hours after reducing the MEP pathway flux by treatment with specific inhibitors ( Fig 5 ) . We therefore conclude that stress situations ( including those causing a decreased DXS activity and/or MEP pathway flux ) could rapidly trigger an increased accumulation of ClpB3 , but not ClpC chaperones , likely aimed to promote the reactivation pathway that would keep DXS enzymes in an enzymatically active condition . Furthermore , our data show that ClpB3 levels are more prone to change compared to those of ClpC proteins , suggesting that ClpB3 concentration might be a major factor regulating the fate of inactive DXS polypeptides recognized by J20 and delivered to Hsp70 . Based on the presented data , we propose a model for the regulation of DXS enzyme levels and activity by different types of plastidial chaperones ( Fig 6 ) . According to this model , J20 ( a plastidial member of the J-domain protein family , also known as J-proteins or Hsp40 co-chaperones ) acts as an adaptor providing substrate specificity [19] . In particular , J20 delivers inactive DXS proteins to Hsp70 chaperones that would next act together with particular Hsp100 proteins to either degrade ( ClpC1 ) or reactivate ( ClpB3 ) the enzyme ( Fig 6A ) . J20 might recognize DXS polypeptides that remain unfolded after plastid import or become misfolded by ordinary perturbations and eventually aggregate ( S3 and S4 Figs ) , a process that would render the protein more insoluble and enzymatically inactive . Under normal growth conditions , most DXS proteins remain soluble but some are indeed found associated to the insoluble fraction ( Fig 6B ) . This might be due to the relative low levels of ClpB3 relative to ClpC1 ( S7 Fig ) [42 , 67] . In agreement , a further reduction in ClpB3 levels ( e . g . in the clpb3 mutant ) results in a higher proportion of DXS protein associated to the insoluble fraction ( Fig 6B ) and hence inactive ( Fig 4 ) . By contrast , an enhanced accumulation of ClpB3 takes place in stress situations ( Fig 5 and S7 Fig ) or when Clp protease function is impaired ( Figs 1B and 3A and S2 Fig ) [21 , 22 , 24 , 40 , 65] , likely aimed to mitigate general protein folding stress . In the case of DXS , a reduced degradation rate in the clpc1 mutant ( Fig 2A ) results in increased levels of active ( soluble ) enzyme ( Figs 3 and 6B ) likely because a higher accumulation of ClpB3 prevents DXS aggregation . Similar to that proposed in other systems [47–50 , 61–64] , ClpB3 directly interacts with Hsp70 to synergistically perform this role ( Fig 4 ) . The observed changes in DXS protein levels and solubility appear to be highly specific , as the next enzyme of the MEP pathway ( Fig 5A ) , deoxyxylulose 5-phosphate reductoisomerase ( DXR ) , was found to be essentially soluble in WT and Hsp100-defective mutants ( Fig 6B ) and to remain unchanged in J20-defective plants [19] . Interaction with CHIP , a co-chaperone that functions as an E3 ubiquitin ligase , converts Hsp70 from a protein-folding machine into a degradation factor that targets unfolded substrates for degradation by the eukaryotic 26S proteosome [51 , 71] . Based on genetic evidence ( Fig 3 ) and published results that ClpC and Hsp70 chaperones can be found together in plastid complexes [54 , 55] , we propose that Hsp70 and ClpC chaperones could somehow interact ( either directly or by means of unidentified partners ) to deliver client proteins like DXS to the Clp catalytic complex . In summary , our model ( Fig 6A ) proposes that collaboration of Hsp70 with Hsp100 chaperones might deliver inactive ( misfolded or/and aggregated ) forms of DXS ( and potentially many other plastidial proteins recognized by specific J-proteins , the substrate adaptors for Hsp70 ) to either refolding ( via ClpB3 ) or degradation ( via ClpC chaperones ) . The seedling lethal phenotype of double mutants with no ClpB3 and Clp protease activity [22] illustrates the key relevance of these two seemingly antagonistic pathways for plant life . We speculate that taking a specific pathway ( i . e . deciding whether to repair or degrade the protein ) might depend on the relative abundance of these Hsp100 partners , particularly as a consequence of changes in ClpB3 levels . The main reason behind the existence of such sophisticated and expensive pathways for the regulation of DXS levels and activity is likely to be the major role demonstrated for this enzyme in the control of the MEP pathway flux [10 , 11 , 13] . Future work should next determine how the collaboration of different sets of plastidial chaperone types , and hence the fate of the client protein , is specifically regulated .
Arabidopsis thaliana mutant lines used here are indicated in S1 Table ( all in the Columbia background ) . Sibling lines expressing 35S:DXS-GFP in WT and j20 backgrounds were previously generated [19] . Seeds were surface-sterilized and germinated on solid Murashige and Skoog ( MS ) medium supplemented with 1% sucrose . Plants were grown under long day conditions as described [19] . For cycloheximide experiments , seeds were germinated on top of a sterile disc of synthetic fabric ( SefarNitex 03-100/44 ) . At day 7 , the disc with the seedlings was transferred to fresh medium supplemented with 100 μM cycloheximide and samples were collected at different times afterwards ( up to 12h ) for immunoblot analysis . Inhibition of protein synthesis with cycloheximide had no visual effect on treated seedlings at the times used for the experiment ( S8 Fig ) . Treatments with MEP pathway inhibitors were performed by transferring discs with 7-day-old seedlings to fresh medium supplemented with 10 μM clomazone ( CLM ) or 100 μM fosmidomycin ( FSM ) . CLM resistance was estimated by quantifying chlorophyll levels in the presence of increasing concentrations of the inhibitor as described [60] . For transient expression and co-immunoprecipitation assays , an Arabidopsis full-length cDNA encoding ClpC1 without the stop codon was PCR-amplified , cloned into the pDONOR207 vector ( Invitrogen ) , and subcloned into the Gateway vector pGWB417 to be expressed under the 35S promoter with a C-terminal MYC epitope ( 35S:ClpC1-MYC construct ) . A 35S:DXS-GFP construct was available in the lab [19] . Transient expression of these constructs was carried out by agroinfiltration of Nicotiana benthamiana leaves using the Agrobacterium GV3101 strain . Samples for immunoprecipitation were collected after 3 days . Protein extracts were obtained from whole plants and used for immunoprecipitation assays or/and immunoblot analysis as described [19] . For the separation of soluble and insoluble ( with protein aggregates ) fractions , native protein extracts were obtained in a buffer containing 100 mM Tris-HCl pH7 . 9 , 10 mM MgCl2 , 1% ( v/v ) glycerol , and 20 μl/ml protease inhibitor cocktail ( Sigma ) . After centrifugation for 10 min at 10 . 000 xg , the supernatant was collected as the soluble fraction . The pellet was washed with fresh buffer and centrifuged again . The obtained pellet fraction was then resuspended in denaturing TKMES buffer [19] and centrifuged again to collect the supernatant as the insoluble fraction . In all cases , protein concentration was determined using the Bio-Rad protein assay . For immunoblot assays , antibodies raised against DXS and DXR [19] , GFP ( Life Technologies ) , MYC ( Millipore ) , and chloroplast Hsp70 , ClpC , and ClpB proteins ( Agrisera ) were diluted 1:500 for DXS , 1:7 , 000 for DXR , 1:1 , 000 for GFP and MYC , 1:6 , 000 for Hsp70 , 1:2 , 000 for ClpC , and 1:3 , 000 for ClpB . The total amount of protein loaded per lane was calculated for each particular antibody to remain in the linear range ( S9 Fig ) . Chemiluminescent signals were visualized using a LAS-4000 ( Fujifilm ) image analyzer and quantified with Quantity One ( Bio-Rad ) . Student´s t test was used to assess statistical significance of quantified differences . For protease accessibility assays , protein extracts from 10-day-old WT and j20 seedlings containing 30 μg of total protein were incubated for 5 min at 37°C with increasing concentrations of Proteinase K ( Invitrogen ) . After stopping the reaction with SDS-PAGE loading buffer , extracts were used for immunoblot analysis . DXS enzyme activity measurements were carried out as described [19 , 72] . Specific activity was calculated by dividing the total activity measured in extracts with the amount of DXS protein found in the corresponding sample . RNA isolation , cDNA synthesis , and qPCR experiments were performed as described [19] using the APT1 ( At1g27450 ) gene for normalization . The Aggrescan3D algorithm [73] was used to analyze protein aggregation propensity . Predictions were performed in static mode using a distance of aggregation analysis of 10 Å . The Arabidopsis DXS structure was modelled using Swiss-Model [74] on top of the 2 . 40 Å resolution E . coli DXS structure with PDB code 2O1S . Residues 72 to 707 of the Arabidopsis DXS monomer , sharing a sequence identity of 41 . 08% with the E . coli protein , were structurally aligned and modelled . The interface of the generated homodimer was evaluated with PDBePISA ( http://www . ebi . ac . uk/pdbe/pisa/ ) rendering an area of 8892 Å and a predicted dissociation ΔG for the dimer of 51 . 2 kcal/mol ( close to those of the template E . coli crystal structure , which exhibits an interface of 7970 Å and a dissociation ΔG of 59 . 1 kcal/mol ) .
|
In this paper we report a relatively simple mechanism by which plant chloroplasts deal with inactive forms of DXS , the main rate-determining enzyme for the production of plastidial isoprenoids relevant for photosynthesis and development . We provide evidence supporting that particular members of the Hsp100 chaperone family contribute to either refold or degrade inactive DXS proteins specifically recognized by the J-protein adaptor J20 and delivered to Hsp70 chaperones . Our results also unveil a J-protein-based mechanism for substrate delivery to the Clp complex , the main protease in the chloroplast stroma . Together , this work allows a better understanding of how chloroplasts get rid of damaged DXS ( and potentially other proteins ) , which should contribute to take more informed decisions in future approaches aimed to manipulate the levels of plastidial metabolites of interest ( including vitamins , biofuels , or drugs against cancer and malaria ) in crop plants .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plastids",
"molecular",
"probe",
"techniques",
"enzymes",
"plant",
"cell",
"biology",
"brassica",
"enzymology",
"chloroplasts",
"plant",
"science",
"model",
"organisms",
"chaperone",
"proteins",
"molecular",
"biology",
"techniques",
"plants",
"cellular",
"structures",
"and",
"organelles",
"enzyme",
"chemistry",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"immunoblot",
"analysis",
"proteins",
"enzyme",
"regulation",
"molecular",
"biology",
"biochemistry",
"plant",
"cells",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"proteases",
"cellular",
"types",
"organisms"
] |
2016
|
Specific Hsp100 Chaperones Determine the Fate of the First Enzyme of the Plastidial Isoprenoid Pathway for Either Refolding or Degradation by the Stromal Clp Protease in Arabidopsis
|
Emergency myelopoiesis is inflammation-induced hematopoiesis to replenish myeloid cells in the periphery , which is critical to control the infection with pathogens . Previously , pro-inflammatory cytokines such as interferon ( IFN ) -α and IFN-γ were demonstrated to play a critical role in the expansion of hematopoietic stem cells ( HSCs ) and myeloid progenitors , leading to production of mature myeloid cells , although their inhibitory effects on hematopoiesis were also reported . Therefore , the molecular mechanism of emergency myelopoiesis during infection remains incompletely understood . Here , we clarify that one of the interleukin ( IL ) -6/IL-12 family cytokines , IL-27 , plays an important role in the emergency myelopoiesis . Among various types of hematopoietic cells in bone marrow , IL-27 predominantly and continuously promoted the expansion of only Lineage−Sca-1+c-Kit+ ( LSK ) cells , especially long-term repopulating HSCs and myeloid-restricted progenitor cells with long-term repopulating activity , and the differentiation into myeloid progenitors in synergy with stem cell factor . These progenitors expressed myeloid transcription factors such as Spi1 , Gfi1 , and Cebpa/b through activation of signal transducer and activator of transcription 1 and 3 , and had enhanced potential to differentiate into migratory dendritic cells ( DCs ) , neutrophils , and mast cells , and less so into macrophages , and basophils , but not into plasmacytoid DCs , conventional DCs , T cells , and B cells . Among various cytokines , IL-27 in synergy with the stem cell factor had the strongest ability to augment the expansion of LSK cells and their differentiation into myeloid progenitors retaining the LSK phenotype over a long period of time . The experiments using mice deficient for one of IL-27 receptor subunits , WSX-1 , and IFN-γ revealed that the blood stage of malaria infection enhanced IL-27 expression through IFN-γ production , and the IL-27 then promoted the expansion of LSK cells , differentiating and mobilizing them into spleen , resulting in enhanced production of neutrophils to control the infection . Thus , IL-27 is one of the limited unique cytokines directly acting on HSCs to promote differentiation into myeloid progenitors during emergency myelopoiesis .
Emergency myelopoiesis is inflammation-induced hematopoiesis , which is critical for controlling systemic infection with pathogens such as a virus , bacteria , or parasite [1 , 2] . In contrast to adaptive immune cells such as T cells and B cells , which can vigorously proliferate in response to their specific antigens , innate immune cells need to be replenished from hematopoietic stem cells ( HSCs ) and progenitors in bone marrow ( BM ) because of their low proliferative activity . However , the molecular mechanism of emergency myelopoiesis during infection remains incompletely understood . HSCs and hematopoietic progenitors can directly sense the presence of pathogens via pattern recognition receptors ( Rs ) such as Toll-like receptors ( TLRs ) , and they can also respond to pro-inflammatory cytokines such as interferon ( IFN ) -α , IFN-γ , interleukin ( IL ) -1 , tumor necrosis factor ( TNF ) -α , and granulocyte colony-stimulating factor ( G-CSF ) [1] . IFN-α and IFN-γ have pleiotropic effects on many cell types , including HSCs and hematopoietic progenitors [1] . Recently , these cytokines were demonstrated to induce an expansion of HSCs and myeloid progenitors , leading to the production of mature myeloid cells [3–6] , although their inhibitory effects on hematopoiesis were previously reported [7–9] . Currently , thus , there are several conflicting positive and negative effects of IFN-α and IFN-γ in hematopoiesis [10 , 11] . However , these discrepancies may be explained by compensatory mechanisms , including IFN-γ-mediated secretion of other cytokines such as IL-6 [12] and fms-related tyrosine kinase 3 ligand ( Flt3L ) [13] . IL-27 is one of the IL-6/IL-12 family cytokines; it plays important roles in immune regulation with both pro-inflammatory and anti-inflammatory properties [14–16] . IL-27 consists of p28 and Epstein-Barr virus-induced gene 3 ( EBI3 ) , and its receptor is composed of WSX-1 and glycoprotein ( gp ) 130 , which is a common receptor subunit in many of the IL-6 family cytokines . We previously demonstrated that IL-27 plays a role in HSC regulation , and that IL-27 expands HSCs and promotes their differentiation in vitro [17] . Moreover , transgenic ( Tg ) mice expressing IL-27 showed enhanced myelopoiesis in BM and extramedullary hematopoiesis in the spleen [17] . In the present study , we further examined the effects of IL-27 on hematopoiesis , the molecular mechanisms , and the physiological role of IL-27 in the control of malaria infection . IL-27 acted on and expanded Lineage ( Lin ) −Sca-1+c-Kit+ ( LSK ) cells , which are highly enriched in HSCs together with very primitive hematopoietic progenitors [18 , 19] , in BM cells in synergy with stem cell factor ( SCF , c-Kit ligand ) and differentiated HSCs into myeloid progenitors through activation of signal transducer and activator of transcription 1 ( STAT1 ) and STAT3 . Moreover , malaria infection induced IFN-γ production , which augmented IL-27 expression , and the IL-27 then promoted the expansion and mobilization of LSK cells into the spleen , resulting in enhanced myelopoiesis to resolve the infection . Our results revealed that IL-27 is one of the limited unique cytokines directly acting on long-term HSCs ( LT-HSC ) , which represent the true stem cells capable of self-renewing , and promotes the expansion and differentiation of them into myeloid progenitors .
Previously , we demonstrated that stimulation of LSK cells with IL-27 and SCF induces an expansion of HSCs and hematopoietic progenitors , including short-term repopulating cells [17] . Moreover , we found that only the combination of IL-27 and SCF , but not either alone , vigorously and continuously expands BM cells to produce LSK cells and CD11b+c-Kit− cells [17] . To examine which cell populations in BM cells respond to IL-27 and SCF in more detail , BM cells were divided into two populations positive or negative for Lin markers except CD11b , and the Lin− population was further divided into four populations positive for either c-Kit or CD11b , or both positive , or both negative . Each population purified by sorting was then stimulated with IL-27 and SCF . Among these five populations , only the Lin−c-Kit+ population greatly expanded ( Fig 1A ) . Next , the BM cells were divided into respective hematopoietic progenitors according to the expression of cell surface markers , as reported previously [20–22] , and stimulated with IL-27 and SCF . Only the LSK cell population vigorously and continuously expanded over more than 6 weeks , although transient and slight expansion was seen in the cell populations of granulocyte/macrophage progenitor ( GMP ) , common myeloid progenitor ( CMP ) , and megakaryocyte/erythrocyte progenitor ( MEP ) ( Fig 1B–1E ) . The expanding cells in the LSK cell population were further analyzed for the expression of cell surface markers . In line with the preliminary results , there seemed to be two populations , the phenotypical LSK population and the Lin+c-Kit− population ( Fig 1E ) . We previously demonstrated that IL-27 Tg mice , which express high amounts of IL-27 in blood , show an increased number of LSK cells in the BM and spleen [17] . To further examine the in vivo effects of IL-27 on the expansion of LSK cells , LSK cells were purified by sorting from BM cells of GFP Tg mice and transferred into wild-type ( WT ) and IL-27 Tg mice . The transferred GFP+ LSK cells vigorously expanded in the BM and spleen of IL-27 Tg mice , but not in those of WT mice , and approximately half of the expanding cells retained the cell surface markers for LSK phenotype ( Fig 1F ) . These results suggest that IL-27 vigorously and continuously augments the expansion of LSK cells both in vitro and in vivo . Because it was previously reported that IFN-α and IFN-γ induce proliferation of HSCs in vivo [3–5] , we next explored the effects of various cytokines in collaboration with SCF on the expansion of LSK cells in vitro . However , IFN-α and IFN-γ augmented the expansion of LSK cells very little , and only IL-27 enhanced it vigorously over 4 weeks ( Fig 1G and 1H ) . Moreover , although there are several cytokines , such as IL-3 , IL-11 , G-CSF , and TPO , that are known to transiently expand and differentiate HSCs [1] , none showed an ability superior to that of IL-27 in expanding LSK cells retaining the LSK phenotype over a long period of time ( S1 Fig ) . Thus , IL-27 has the strongest ability to augment the expansion of LSK cells . The LSK cells expanded by IL-27 and SCF were further analyzed for the cell surface expression of various markers , and the expression levels were compared with those of primary LSK cells freshly prepared from BM of WT mice . The expression levels of macrophage colony-stimulating factor receptor ( M-CSFR ) , CD16/32 , and MHC class II in the LSK cells expanded by IL-27 and SCF were much higher than those in primary LSK cells ( Fig 2A ) . The expression levels of CD34 and CD150 in the expanded LSK cells were slightly less than those in primary LSK cells ( Fig 2A ) . In contrast , the expanded LSK cells were almost completely negative for Flt3 expression , whereas primary LSK cells were positive for Flt3 ( Fig 2A ) . Thus , IL-27 and SCF expand and differentiate primary LSK cells into M-CSFR+Flt3−CD16/32+ LSK cells ( myeloid progenitor cells ) . Next , multipotency of the LSK cells expanded by IL-27 and SCF were examined under various differentiation conditions for migratory dendritic cells ( mDCs ) by granulocyte/macrophage ( GM ) -CSF , plasmacytoid DCs ( pDCs ) and conventional DCs ( cDCs; lymphoid-resident DCs ) were examined using Flt3L and thrombopoietin ( TPO ) [23] , myeloid cells were examined using IL-3 and SCF , and T cells and B cells were examined by using thymic stromal cells ( TSt4 ) with and without expressing Notch ligand Delta-like 1 ( DLL1 ) , respectively [24] . The expanded LSK cells much more rapidly proliferated and differentiated into MHC class II+CD11c+ mDCs than primary LSK cells in response to GM-CSF , although the total number of mDCs achieved seemed to be similar for both ( Fig 2B and 2C ) . However , LSK cells stimulated with IL-27 and SCF rapidly lost the ability to differentiate into pDC and cDC ( Fig 2D and 2E ) . These phenomena are highly consistent with the almost complete abolishment of Flt3 expression on the expanded LSK cells ( Fig 2A ) . Under myeloid differentiation conditions , the expanded LSK cells differentiated much more greatly into neutrophils and slightly into mast cells , but less so into macrophages and basophils ( Fig 2F and 2G ) . Similar enhanced differentiation into myeloid cells was observed using the LSK cells obtained from WT and IL-27 Tg mice ( S2 Fig ) . In contrast , the ability to differentiate into B cells and T cells was almost completely abrogated in the expanded LSK cells ( Fig 2H ) . Moreover , the ability of the LSK cells expanded by IL-27 and SCF to differentiate into myeloid cells in vivo was explored by using mixed BM chimeras . The equal cell numbers of the LSK cell population expanded from CD45 . 1 congenic mice and BM cells from CD45 . 2 congenic mice were mixed and transferred into sublethally irradiated CD45 . 2 congenic mice . After 9 days , cell populations of neutrophils in the BM and spleen were analyzed . In agreement with the in vitro results described , the percentages of neutrophils derived from the expanded LSK cells were markedly higher than those from primary LSK cells in both BM and spleen ( Fig 2I ) . To explore the molecular mechanisms whereby IL-27 and SCF expand LSK cells , total RNA was prepared from the expanded LSK cells and respective hematopoietic progenitors and analyzed with real-time reverse transcriptase-polymerase chain reaction ( RT-PCR ) . The expanded cells highly expressed the transcription factors critical for differentiation into myeloid cells such as Spi1 , Gfi1 , and Cebpa , but expression was much less for those important for the other types of cells such as Irf8 , Tcf4 , and Ikzf1 [25–27] ( Fig 2J ) . No expression of transcription factors important for B cells and erythrocytes , Pax5 [28] and Gata1 [29] , respectively , was observed . In addition , the expression of Cbepb , which is a transcription factor recently demonstrated to be regulated by cytokines and control emergency granulopoiesis [30–32] , was also increased ( S3 Fig ) . These results suggest that LSK cells expanded by IL-27 and SCF are multipotent myeloid progenitors that have unique potential to differentiate into mDCs , neutrophils , and mast cells , and less so into macrophages , and basophils , but not into pDCs , cDCs , T cells , and B cells . We and others previously demonstrated that IL-27 activates both STAT1 and STAT3 through WSX-1 and gp130 , respectively [14 , 15] . Consistent with these reports , real-time RT-PCR analysis revealed that LSK populations purified from LSK cells expanded by IL-27 and SCF and primary BM cells were positive for mRNA expression of STAT1 and STAT3 ( Fig 3A ) . Phosphorylation of STAT1 and STAT3 was also detected in primary WT LSK cells , but not WSX-1-deficient LSK cells ( Fig 3B ) , in response to IL-27 and SCF , which were analyzed by flow cytometry . Furthermore , IL-27 alone induced phosphorylation of both STAT1 and STAT3 , whereas SCF alone failed to induce phosphorylation of either one , as discussed previously [33] ( S4 Fig ) . To further investigate the roles of STAT1 and STAT3 in the expansion of LSK cells and the ability to differentiate into myeloid progenitors by IL-27 and SCF , we used STAT1-deficient LSK cells and conditional STAT3-knockout ( STAT3 cKO ) LSK cells . LSK cells from WT ( 129 ) and STAT1-deficient mice were stimulated with IL-27 and SCF . Although WT LSK cells comprised more than 90% of cells with the LSK phenotype after 7 days , STAT1-deficient LSK cells comprised half LSK phenotype cells and half Sca-1−LK ( LS−K ) ( Fig 3C ) . Nevertheless , the number of expanded cells was comparable among them , probably due to the anti-proliferative effects by STAT1 signaling [34] . However , the purified STAT1-deficient LS−K cells failed to survive thereafter , even in the presence of IL-27 and SCF ( Fig 3D ) , although the purified STAT1-deficient LSK cells expanded well ( Fig 3D ) , as did WT 129 LSK cells ( Fig 3C ) . Moreover , WT and STAT1-deficient LSK cells had similar abilities to differentiate into MHC class II+CD11c+ mDC cells ( Fig 3E ) and macrophages ( Fig 3F ) . STAT1-deficient LSK cells showed reduced ability to differentiate into neutrophils and mast cells , but increased ability to differentiate into basophils ( Fig 3F ) . In line with the reduced ability to differentiate into neutrophils , mRNA expression of the critical transcription factor Gfi1 was significantly reduced in STAT1-deficient LSK cells compared with that in WT LSK cells ( Fig 3G ) . In contrast , STAT3 cKO LSK cells expanded very little in response to IL-27 and SCF ( Fig 3H ) . Moreover , the residual surviving LSK cells almost completely lost the ability to differentiate into mDCs ( Fig 3I ) and myeloid cells ( Fig 3J ) . Consistent with the abrogated abilities , these cells showed reduced expression of the critical transcription factors such as Spi1 , Gfi1 , and Cebpa , but not Irf8 ( Fig 3K ) . These results suggest that both STAT1 and STAT3 are necessary for LSK cells to fully expand and differentiate into myeloid progenitor cells in response to IL-27 and SCF . To more precisely define which cell population responds to stimulation with IL-27 and SCF , LSK cells were further divided into two populations according to CD34 expression . Although the percentage of the more primitive population of CD34− LSK cells was much less than that of CD34+ LSK cells , the CD34− LSK cells responded much better to stimulation with IL-27 and SCF and expanded more vigorously than CD34+ LSK cells ( Fig 4A and 4B ) . Then , the LSK cells were further divided into eight populations , F1 to F8 , including LT-HSCs ( CD34−CD150+CD41− LSK , F1 ) and myeloid-restricted progenitor cells with long-term repopulating activity ( MyRPs , CD34−CD150+CD41+ LSK , F4 ) according to the recently revised criteria [19] . Respective populations purified by sorting ( Fig 4C ) were stimulated with IL-27 and SCF . Only two populations , F1 and F4 , vigorously expanded . F5 , which corresponds to populations more differentiated toward myeloid cells such as macrophages , slightly expanded ( Fig 4D and 4E ) . The F1 and F4 populations expanded by IL-27 and SCF had great abilities to differentiate into myeloid cells , particularly neutrophils ( Fig 4E ) . Thus , IL-27 and SCF expand CD34−CD150+ LSK cells , including LT-HSCs and MyRPs , and differentiate them into myeloid progenitor cells , which have great potential to differentiate mainly into neutrophils . We previously demonstrated that in the blood stage of malaria infection with the attenuated variant Plasmodium ( P ) berghei XAT derived from the lethal strain P . berghei NK65 IFN-γ production induced by IL-12 and phagocytic cells in the spleen are critical for controlling parasitemia [35 , 36] . Recently , it was reported that the blood stage of P . chabaudi infection induces mobilization of early myeloid progenitor cells out of BM , thereby transiently establishing myelopoiesis in the spleen through IFN-γ to resolve the infection [6 , 37] . In line with these results , WSX-1-deficient mice showed more increased parasitemia than WT mice at 7 days ( Fig 5A ) , just prior to when the parasitemia reaches its peak after infection with P . berghei XAT ( S5A Fig ) . In contrast , no significant difference was observed in the serum IFN-γ levels in WT and WSX-1-deficient mice ( Fig 5B ) . The infection markedly induced the enhanced percentage and number of LSK cells in the BM and spleen of WT mice ( Fig 5C and 5D ) . The LSK cells in the BM showed greatly augmented abilities to differentiate in vitro into neutrophils and mast cells , but had slightly reduced abilities to differentiate into macrophages and basophils after infection ( Fig 5E ) . Moreover , the LSK cells in the spleen exhibited a much more enhanced ability to differentiate into neutrophils , macrophages , and mast cells ( Fig 5E ) . In contrast , of note , WSX-1-deficient mice showed significantly reduced percentage and number of LSK cells in the BM and spleen compared with WT mice after infection ( Fig 5C and 5D ) . In particular , the cell number of the neutrophils in the spleen was increased very little in WSX-1-deficient mice ( Fig 5F and 5G ) . Consistent with this , LSK cells from BM of WSX-1-deficient mice showed reduced abilities to differentiate in vitro into neutrophils after infection compared with those of WT mice ( S6 Fig ) . In addition , more greatly reduced abilities to differentiate in vitro into various myeloid cells were observed when LSK cells from spleen were used ( S6 Fig ) . Moreover , mixed bone marrow chimera experiments using bone marrow cells from WT and WSX-1-deficient mice revealed that the effect of IL-27 on the expansion of LSK cells and neutrophils is actually a cell-autonomous direct effect ( S7 Fig ) . To further elucidate the protective role of LSK cells in malaria infection , LSK cells purified from BM cells of infected WT CD45 . 1 mice were injected into the infected WSX-1-deficient CD45 . 2 mice . Consistent with the increased percentage of neutrophils differentiated from the WT LSK cells in the BM and spleen of transferred WSX-1-deficient mice ( Fig 5H ) , parasitemia was significantly decreased in the WSX-1-deficient mice by the transfer of WT LSK cells compared with that in non-transferred WSX-1-deficient mice ( Fig 5I ) . Thus , the blood stage of malaria infection induces expansion , differentiation , and mobilization of LSK cells into the spleen to produce myeloid cells such as neutrophils in an IL-27-dependent manner . Although it was previously reported that IFN-α and IFN-γ induce proliferation of HSCs in vivo [3–5] , IFN-α and IFN-γ augmented the expansion of LSK cells very little in vitro , and only IL-27 enhanced it vigorously over 4 weeks ( Fig 1G and 1H ) . However , similar to the work previously reported [37 , 38] , IFN-γ-deficient mice showed increased parasitemia with almost no increase in the number of LSK cells in BM and spleen ( Fig 6A and 6B ) . To clarify the molecular mechanism whereby IFN-γ induces the expansion of LSK cells , the expression of IL-27 subunits EBI3 and p28 was examined . Although the infection did not increase EBI3 mRNA expression in the BM and spleen of both WT and IFN-γ-deficient mice ( S8 Fig ) , intriguingly , the infection greatly enhanced p28 mRNA expression in WT mice but failed to enhance it in IFN-γ-deficient mice ( Fig 6C ) . In agreement with this , p28 protein levels in the serum were greatly increased by the infection in WT mice but not in IFN-γ-deficient mice ( Fig 6D ) . Consistent with the in vivo role of IFN-γ , we also observed the augmentation of mRNA expression of p28 , but not EBI3 , and p28 protein production in the culture supernatants of WT BM cells stimulated with IFN-γ in vitro ( S9 Fig ) . To further clarify the role of IL-27 downstream of IFN-γ , we next performed the experiment to see the effects of forced expression of IL-27 on the susceptibility to malaria infection in IFN-γ-deficient mice . The hydrodynamic injection of IL-27 expression vector into the infected IFN-γ-deficient mice showed significantly decreased parasitemia compared with that of control vector ( Fig 6E ) . This phenomenon was accompanied by the enhanced percentage of LSK cells and augmented numbers of LSK cells and neutrophils in both BM and spleen ( Fig 6F and 6G ) . Thus , the blood stage of malaria infection augments the expression of IL-27 through IFN-γ , and IL-27 then promotes the expansion , differentiation , and mobilization of LSK cells into the spleen to control parasitemia .
Previously , we found that IL-27 , which is in the IL-6/IL-12 family of cytokines , plays a role in the regulation of HSCs in vitro and in vivo [17] . Here , we have further elucidated that IL-27 is a unique cytokine that directly acts on LSK cells to promote their differentiation into myeloid progenitor cells called M-CSFR+Flt3−CD16/32+ LSK cells , which still retain the LSK phenotype ( Fig 2A–2I ) . These progenitors have great potential to give rise to neutrophils , mDCs , and mast cells , but not to pDCs , cDCs , T cells , and B cells . Interestingly , among various BM progenitor cells , IL-27 and SCF vigorously and continuously expand only HSCs and primitive myeloid progenitor cells with long-term repopulating activity , such as LT-HSCs and MyRPs , respectively [19] , for more than 4 weeks ( Fig 4 ) . Consistent with the ability to differentiate into myeloid progenitor cells , the LSK cells expanded by IL-27 and SCF expressed transcription factors such as Spi1 , Gfi1 , Cebpa , and Cebpb , which are critical for myeloid differentiation [25–27 , 30–32] ( Fig 2J and S3 Fig ) . Although Cebpb was reported to be an important transcription factor for emergency granulopoiesis [30–32] , STAT3 signaling was revealed to be important for its upregulation , whereas STAT1 signaling unexpectedly suppressed its expression ( S10 Fig ) . This phenomenon seems to correlate to the expression level of the anti-apoptotic gene Bcl-2 [39 , 40] , but not the transcription factor E2-2 , which is critical for pDC differentiation [41] . Further studies are necessary to elucidate the precise roles of each STAT in the regulation of Cebpb expression . Thus , IL-27 is one of the limited unique cytokines that directly acts on the most primitive LT-HSCs; it promotes their expansion and differentiation into myeloid progenitor cells , presumably through MyRPs [19] , to replenish myeloid cells such as neutrophils in the periphery during emergency myelopoiesis . Sca-1 is an IFN-responsive molecule that is highly upregulated in many hematopoietic cells following exposure to IFNs [10 , 11 , 37] . Consequently , myeloid-restricted progenitor cells normally identified as Lin−Sca-1−c-Kit+ ( LS−K ) become positive for Sca-1 and can no longer be distinguished from the real multipotent LSK cells , resulting in overestimation of the latter population , and this is a problem . IL-27 was previously reported to enhance the expression of Sca-1 on T cells [42] . However , to alleviate the problem , we initially identified the cell population responsive to IL-27 and SCF among BM cells by using LSK cells and various hematopoietic progenitor cells purified by sorting . Intriguingly , it turned out that the SCF and LSK cell populations expanded vigorously and continuously in response to IL-27 , and that the LS−K cell populations including GMP , CMP , and MEP only transiently and slightly responded during the first week and then disappeared thereafter ( Fig 1A–1E ) . Moreover , in almost all in vitro experiments , we used primary LSK cells that were purified by sorting . The high responsiveness of the LSK cells to IL-27 seems to be partially due to the higher mRNA expression of WSX-1 in the LSK cells compared to that of other hematopoietic progenitor cells ( S3 Fig ) [17] . We previously demonstrated that during the blood stage of malaria infection with attenuated P . berghei XAT , IL-12-mediated IFN-γ production and phagocytic cells ( including neutrophils ) in the spleen are critical for controlling parasitemia [35 , 36 , 43] . Previous studies demonstrated that neutrophils play an important role in killing malaria parasites in mice , rats , and humans [43–45] . In marked contrast , regarding infection with lethal P . berghei NK65 , IL-12-mediated IFN-γ production was shown to contribute to T-cell-dependent immunopathology [46] . However , a major role of IL-27 in infection is its suppression of excess immune responses against infection by controlling the production of pro-inflammatory cytokines [14–16] . Consistent with this , WSX-1/IL-27 was recently demonstrated to have a critical role in limiting the effector CD4+ T-cell-mediated immunopathology caused by IL-12-dependent IFN-γ production during infection with lethal P . berghei NK65 [47–49] . The present study clearly revealed that WSX-1/IL-27 contributes to clearance of parasites due to enhanced myelopoiesis during the early phase of infection with attenuated P . berghei XAT ( Fig 5 and S5A Fig ) . However , during the late phase of infection , WSX-1/IL-27 seems to play a role in limiting the production of pro-inflammatory cytokines such as IFN-γ ( S5B Fig ) , leading to augmented reduction of parasitemia ( S5A Fig ) , as in the case of infection with lethal P . berghei NK65 [47–49] . Moreover , our preliminary data revealed that there were no apparent differences observed in parasitemia or expansion of LSK cells and neutrophils in the BM and spleen of WT and WSX-1-deficient mice 7 days after infection with lethal P . berghei NK65 ( S11 Fig ) . It is conceivable that pro-inflammatory cytokines other than IL-27 were abundantly produced in the absence of IL-27 during infection with lethal P . berghei NK65 and that late-phase infection with attenuated P . berghei XAT may have redundantly compensated for the loss of IL-27 to promote myelopoiesis . However , other studies have shown that IL-27 limits migration of neutrophils from the BM to the site of inflammation by reducing production of cytokines and chemokines during influenza infection [50] and septic peritonitis [51] . IL-27 was also reported to be a negative regulator of neutrophil function [52] . Although IL-27 directly promotes myelopoiesis to produce myeloid progenitors in BM , as shown in the present study , IL-27 may indirectly regulate migration of these progenitors and neutrophils to the site of inflammation and limit neutrophil function . Thus , IL-27 has both positive and negative effects on neutrophils; therefore , the overall outcome of the effects of IL-27 is likely to be governed by the balance between these effects , depending on the disease model . It was recently demonstrated that P . chabaudi infection induces mobilization of early myeloid progenitor cells out of BM , thereby transiently establishing myelopoiesis in the spleen through IFN-γ [37] . However , the expression of IFN-γR in the hematopoietic compartment was dispensable , whereas its expression in the irradiation-insensitive cellular compartment , including endothelial cells and stromal cells , was important [37] . Secretion of IFN-γ-induced chemokines such as CCL2 and CCL7 by non-hematopoietic cells plays a critical role in the mobilization of CCR2-expressing HSCs [37] . In this study , however , there is no experimental evidence regarding how IFN-γ regulates the activation of HSCs . In the present study , WSX-1-deficient mice showed significantly reduced numbers of LSK cells and neutrophils compared with WT mice after P . berghei XAT infection , resulting in increased parasitemia ( Fig 5A–5F and S5A Fig ) . These results suggest that endogenous IL-27 greatly contributes to the clearance of parasitemia through augmentation of myelopoiesis . Moreover , similar to P . chabaudi infection , P . berghei XAT infection could not increase the number of LSK cells in IFN-γ-deficient mice with increased parasitemia ( Fig 6A and 6B ) . Of note , after P . berghei XAT infection , p28 mRNA expression and its serum protein level were markedly upregulated in an IFN-γ-dependent manner ( Fig 6C and 6D and S9 Fig ) . This is consistent with previous reports indicating that p28 gene transcription in macrophages is induced by IFN-γ and TLR ligands [53] , and that IFN-γ limits Th17-mediated and Th9-mediated autoimmune inflammation through IL-27 production [54 , 55] . In addition , the hydrodynamic injection of the IL-27 expression vector into infected IFN-γ-deficient mice greatly recovered the number of LSK cells and neutrophils in the BM and spleen and eventually reduced parasitemia ( Fig 6E–6G ) . Thus , during malaria infection , it is highly conceivable that the proliferative effects on LSK cells by IFN-γ are indirectly mediated by IL-27 . In our study , we could not observe any direct proliferative effect of IFN-α and IFN-γ on LSK cells in vitro , as recently pointed out by others [10 , 11] , and only IL-27 augmented the proliferative effect for more than 4 weeks ( Fig 1G and 1H ) . Recently , IL-27 was reported to have a polyglutamic acid domain in the p28 subunit , which is unique among cytokines , and to confer hydroxyapatite-binding and bone-binding properties and bone tropism to bone sialoprotein and the endosteal bone surface [56] . This location in the BM has been identified as a niche for HSCs [57] , and these properties support the idea that IL-27 plays a critical role in the regulation of HSCs in that niche . We detected much higher expressions of IL-27 subunits ( both p28 and EBI3 ) at mRNA levels in BM than in the spleen during the steady state , and P . berghei XAT infection greatly augmented the expression of p28 mRNA in both BM and spleen , and also its serum protein level ( Fig 6C and 6D ) . Further studies are necessary to clarify which BM cells produce IL-27 during malaria infection; mesenchymal stromal cells might be a candidate because of their reported IL-6 production during viral infection [12] , as described in the next section . It was recently demonstrated that specific cytotoxic CD8+ T cells during an acute viral infection with lymphocytic choriomeningitis virus secrete IFN-γ , thus enhancing the production of IL-6 in BM mesenchymal stromal cells and resulting in an increased number of early multipotent progenitors and committed myeloid precursors in the BM and accumulation of myeloid cells in the periphery [12] . The IL-6Rα chain is only expressed at the stage of early multipotent progenitors and downstream myeloid precursors , and it is lacking HSCs [12] . In contrast , IL-27 most predominantly acts on only HSCs , as shown in the present study . Both IL-6 and IL-27 share gp130 , which is ubiquitously expressed as a common receptor subunit . Therefore , downstream of IFN-γ , both IL-27 and IL-6 may be necessary to induce the maximum myelopoiesis to control infection . However , the mode of IL-6 action is complex and there are two major mechanisms: IL-6 classic signaling through membrane IL-6Rα and IL-6 trans-signaling through soluble IL-6Rα [58 , 59] . It was recently reported that IL-6Rα-deficient mice show increased resistance to P . chabaudi infection and that IL-6 trans-signaling , but not IL-6 classic signaling , contributes to a lethal outcome of infection [60] . In contrast to the viral infection , we could not detect any increased mRNA expression of IL-6 in the BM or spleen of WT and IFN-γ-deficient mice with P . berghei XAT infection ( S8 Fig ) . A similar inability of IFN-γ to enhance IL-6 mRNA expression in WT BM cells in vitro was also observed ( S9A Fig ) . In addition , IL-6-deficient mice showed little increased susceptibility to the P . berghei XAT infection , reduced cell numbers in the LSK cell population , and reduced neutrophils in the BM and spleen ( S12 Fig ) . Thus , individual pathogens may utilize different mechanisms to induce emergency myelopoiesis through IL-27 , IL-6 , and others . In conclusion , the present results provide a novel role and mechanism for the action of IL-27 downstream of IFN-γ in the efficient expansion of myeloid progenitor cells from LT-HSCs and MyRP cells and their mobilization into the spleen during acute malaria infection .
The animal study was approved by the Animal Care and Use Committee of Tokyo Medical University ( S-230043 , S-24012 , S-25059 , S-26003 , and S-27009 ) and was performed in accordance with our institutional guidelines and the Fundamental Guidelines for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education , Culture , Sports , Science and Technology , 2006 . C57BL/6 ( CD45 . 2 ) mice and C57BL/6 ( CD45 . 1 ) mice were purchased from Sankyo Lab ( Tokyo , Japan ) . The 129/Sv mice and STAT1-deficient mice ( 129/Sv background ) were purchased from Taconic Farms ( Germantown , NY , USA ) . IL-6-deficient mice ( C57BL/6 background ) were purchased from Jackson Laboratory ( Bar Harbor , ME , USA ) . IFN-γ-deficient mice ( C57BL/6 background ) , STAT3flox/flox mice ( a mixed background of 129/Sv and C57BL/6 ) , and GFP Tg mice ( C57BL/6 background ) [61] were provided by Dr . Iwakura ( Tokyo University of Science ) , Dr . Takeda ( Osaka University ) , Dr . Okabe ( Osaka University ) and Dr . Ito ( Tokyo Medical University ) , respectively . In addition to these mice , IL-27 Tg mice ( C57BL/6 background ) [17] , and WSX-1-deficient mice ( C57BL/6 background ) [62] were maintained in specific pathogen-free conditions under the care of the Laboratory Animal Center of Tokyo Medical University . STAT3 cKO cells were obtained by infecting STAT3flox/flox cells with Cre-expressing retrovirus in vitro . Monoclonal antibodies ( mAbs ) for mouse c-Kit ( 2B8 ) , Sca-1 ( D7 ) , CD3ε ( 145-2C11 ) , CD4 ( GK1 . 5 ) , CD8α ( 53–6 . 7 ) , CD19 ( 6D5 ) , CD49b ( DX5 ) , Gr-1 ( RB6-8C5 ) , TER119/erythroid cell ( TER-119 ) , CD11c ( N418 ) , CD11b ( M1/70 ) , F4/80 ( BM8 ) , NK1 . 1 ( PK136 ) , B220 ( RA3-6B2 ) , FcεRIα ( MAR-1 ) , M-CSFR ( AFS98 ) , Flt3 ( A2F10 ) , CD16/32 ( 2 . 4G2 ) , CD150 ( TC15-12F12 . 2 ) , CD41 ( MWReg30 ) , IL-7Rα ( A7R34 ) , MHC class II I-A/I-E ( M5/114 . 15 . 2 ) , Siglec H ( 551 ) , Ly6G ( 1A8 ) , CD45 . 1 ( A20 ) , and CD45 . 2 ( 104 ) were purchased from BioLegend ( San Diego , CA ) . mAbs against mouse pY701-STAT1 ( 4a ) and pY705-STAT3 ( 4/P-STAT3 ) were purchased from BD Pharmingen ( San Diego , CA ) . mAbs against mouse CD34 ( RAM34 ) was purchased from eBioscience ( La Jolla , CA ) . mAbs against mouse PDCA1 ( JF05-1C2 . 4 . 1 ) was purchased from Miltenyi Biotec ( Bergisch Gladbach , Germany ) . APC-Cy7-conjugated streptavidin , PerCP/Cy5 . 5-conjugated streptavidin , and Brilliant Violet 510-conjugated streptavidin were purchased from BioLegend and used to reveal staining with biotinylated Abs . Mouse recombinant IL-27 and hyper-IL-6 were prepared as a tagged single-chain fusion protein by flexibly linking EBI3 to p28 and soluble IL-6Rα to IL-6 , respectively , using HEK293-F cells ( Life Technologies , Carlsbad , CA ) as described previously [63 , 64] . Mouse recombinant SCF , IL-1β , IL-3 , IL-6 , IL-7 , IL-11 , IL-12 , thymic stromal lymphopoietin ( TSLP ) , G-CSF , M-CSF , GM-CSF , TNF-α , and human recombinant TPO were purchased from PeproTech ( Rocky Hill , NJ ) . Human recombinant Flt3L was purchased from Miltenyi Biotec . Mouse recombinant IL-23 , IL-25 , and IL-33 were purchased from R&D Systems ( Minneapolis , MN ) . Mouse IFN-α was purchased from PBL Biomedical Laboratories ( Piscataway , NJ ) . Mouse recombinant IFN-γ was provided from Shionogi Pharmaceutical Co . , Ltd . ( Osaka , Japan ) , Spleen and BM Lin− cells were enriched by negative selection using an autoMACS Pro ( Miltenyi Biotec ) with a combination of magnetic beads conjugated with mAbs against CD3ε , CD4 , CD8α , Gr-1 , TER119 , CD11b , CD11c , NK1 . 1 , B220 , and FcεRIα . Subsequently , cells were stained with mAbs against c-Kit , and Sca-1 was used for LSK . For multiple fractions of HSC , cells were stained with CD34 , c-Kit , Sca-1 , CD150 , and CD41 mAbs [19] . In the case of common lymphoid progenitor ( CLP ) , macrophage-DC progenitor ( MDP ) , and common DC progenitor ( CDP ) , cells were stained with c-Kit , Sca-1 , IL-7Rα , M-CSFR , and Flt3 mAbs [20–22] . For GMP , CMP , and MEP , cells were stained with CD34 , c-Kit , Sca-1 , and CD16/32 mAbs [20–22] . Sorting was performed on FACS Aria or FACS Aria III ( BD Bioscience ) . Cells were cultured at 37°C under 5% CO2/95% air in RPMI-1640 ( SIGMA , St . Louis , MO ) containing 10% fetal calf serum , 50 μM 2-mercaptoethanol ( GIBCO , Grand Island , NY ) , and 100 μg/ml kanamycin ( Meiji Seika , Tokyo , Japan ) . To proliferate progenitors , sorted cells were cultured with 10 ng/ml IL-27 and/or 10 ng/ml SCF . To examine the effect of various cytokines on proliferation of LSK , IL-1β , IL-3 , IL-6 , hyper-IL-6 , IL-11 , IL-12 , IL-23 , IL-25 , IL-33 , TSLP , G-CSF , TPO , TNF-α , and IFN-α were used as a final concentration of 10 ng/ml . IFN-γ was used as a final concentration of 100 U/ml . Half of the medium was changed every 3 days , with cytokines added . For evaluation of mDC potential , sorted cells ( 1–5 × 103 ) in 96-well plates were cultured with 20 ng/ml GM-CSF for 3 to 10 days . For pDC and cDC potential , sorted cells ( 5 × 103 ) were cultured with 100 ng/ml Flt3L and 20 ng/ml TPO for 10 days . For analysis of multipotent myeloid potential , sorted cells ( 1 × 102–5 × 103 ) in 96-well plates were cultured with 10 ng/ml SCF and 10 ng/ml IL-3 for 6 days . Half of the medium was replaced every 3 days , with cytokines added . For detection of B-cell potential , sorted cells ( 2 × 102 ) were cultured with a monolayer of thymic stromal cells ( TSt4 ) containing 2 ng/ml IL-7 for 15 days [24] . For detection of T-cell potential , sorted cells ( 2 × 102 ) were cultured with TSt4 cells expressing DLL1 , TSt4/DLL1 cells , containing 2 ng/ml IL-7 and 2 ng/ml Flt3L for 18 days [24] . Half of the medium was replaced every 7 days , with cytokines added . The following cell surface markers were used to identify respective cells: mDC; MHC class II+CD11c+ , pDC; Siglec H+PDCA1+CD11c+ , cDC; Siglec H−PDCA1−CD11c+ , neutrophil; Ly6G+CD11b+ or Gr-1+CD11b+ , macrophage; F4/80+CD11b+ , mast cell; c-Kit+FcεR1α+CD11b− , basophil; CD49b+FcεR1α+CD11b− , B cell; B220+CD19+ , double positive T cell; and CD4+CD8+ . Flow cytometry was performed on a FACS Canto II ( BD Bioscience , San Jose , CA ) and data were analyzed using FlowJo Software ( Tree Star , Ashland , OR ) . Cell number was counted using flow cytometry unless otherwise indicated . For intracellular cytokine staining , cells were fixed with Fixation Buffer ( BD Bioscience ) for 30 min and permeabilized with Perm Buffer II ( BD Bioscience ) for 30 min . Then , samples were stained with antibodies for pY701-STAT1 and pY705-STAT3 . Total RNA was prepared using RNeasy Mini Kit ( QIAGEN , Hilden , Germany ) , and cDNA was prepared using oligo ( dT ) primer and SuperScript III RT ( Invitrogen , Carlsbad , CA , USA ) . Real-time quantitative PCR was performed using SYBR Premix Ex Taq II and a Thermal cycler Dice real-time system according to the manufacturer’s instructions ( TAKARA , Otsu , Shiga , Japan ) . Glyceraldehyde-3-phosphate ( GAPDH ) was used as housekeeping gene to normalize mRNA . Relative expression of real-time PCR products was determined by using the ΔΔCt method to compare target gene and GAPDH mRNA expression . Primers used in this study are listed in S1 Table . For in vivo proliferation analysis , BM LSK cells ( 8 × 103 ) purified from GFP Tg mice were intravenously transferred into irradiated ( 4 Gy ) WT and IL-27 Tg mice . To evaluate in vivo development , IL-27/SCF-cultured BM LSK cells ( 2 × 104 ) from CD45 . 1 congenic mice were transplanted into sublethally irradiated ( 6 Gy ) CD45 . 2 recipient mice with the same congenic BM cells ( 2 × 106 ) . As in the case of malaria infection , BM LSK cells ( 5 × 104 ) were sorted from malaria-infected CD45 . 1 mice and intravenously transferred into irradiated ( 4 Gy ) WSX1-deficient mice 7 days before infection . The retroviral vector pMX-Cre-GFP ( from Dr . M . Kubo ) was transfected into Platinum-E packaging cells [65] using FuGENE 6 ( Promega , Madison , WI ) , and supernatants of these cultures were used as the source of viral particles . LSK cells sorted from BM cells of STAT3flox/flox mice were stimulated with IL-27 and SCF ( 10 ng/ml each ) and transduced with viral particles by spin infection ( 2 , 000 rpm , 90 min , 25°C ) using 8 μg/ml Polybrene at 24 hr and 48 hr later . The next day , GFP+ LSK cells were sorted and used as STAT3 cKO LSK cells . Mice were injected intravenously with a red blood cell ( RBC ) suspension containing parasitized RBC ( 1 × 104 ) with the nonlethal strain P . berghei XAT [35] , which is an irradiation-induced attenuated variant of the lethal strain P . berghei NK65 , or the P . berghei NK65 [46] . Parasitemia was assessed by the microscopic examination of Giemsa-stained smears of tail blood after infection . The percentage of parasitemia was calculated as follows: parasitemia ( % ) = [ ( number of infected RBC ) / ( total number of RBC counted ) ] × 100 . IFN-γ-deficient mice were intravenously injected with 25 μg of p3xFLAG-CMV ( Sigma Chemical Co . , St . Louis , MO ) , or p3xFLAG-IL-27 plasmids at days 0 and 4 after malaria infection . Amounts of IL-27 p28 in culture supernatants or serum were determined by using Quantikine kits ( R&D ) according to the manufacturer’s instruction . Data are represented as mean ± SEM . Statistical analyses were performed by two-tailed Student’s t test for two groups , and by one-way ANOVA and Bonferroni’s multiple comparison tests for multiple groups . P < 0 . 05 was considered to indicate a statistically significant difference .
|
Emergency myelopoiesis is inflammation-induced hematopoiesis that is critical for controlling infection with pathogens , but the molecular mechanism remains incompletely understood . Here , we clarify that one of the interleukin ( IL ) -6/IL-12 family cytokines , IL-27 , plays an important role in emergency myelopoiesis . Among various types of hematopoietic cells in bone marrow , IL-27 predominantly and continuously promoted expansion of only Lineage−Sca-1+c-Kit+ ( LSK ) cells , especially long-term repopulating hematopoietic stem cells , and differentiation into myeloid progenitors in synergy with stem cell factor . These progenitors expressed myeloid transcription factors such as Spi1 , Gfi1 , and Cebpa/b through activation of signal transducer and activator of transcription 1 and 3 , and had enhanced potential to differentiate into neutrophils , but not into plasmacytoid dendritic cells . Among various cytokines , IL-27 in synergy with stem cell factor had the strongest ability to augment the expansion of LSK cells and their differentiation into myeloid progenitors . The blood stage of malaria infection was revealed to enhance IL-27 expression through interferon-γ production , and IL-27 then promoted the expansion of LSK cells , differentiating and mobilizing them into the spleen , resulting in enhanced production of neutrophils to control the infection . Thus , IL-27 is one of the limited unique cytokines directly acting on hematopoietic stem cells to promote differentiation into myeloid progenitors during emergency myelopoiesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"plasmodium",
"spleen",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"cell",
"differentiation",
"parasitology",
"parasitemia",
"developmental",
"biology",
"apicomplexa",
"molecular",
"development",
"neutrophils",
"quantitative",
"parasitology",
"white",
"blood",
"cells",
"animal",
"cells",
"immune",
"system",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"malaria"
] |
2016
|
Promotion of Expansion and Differentiation of Hematopoietic Stem Cells by Interleukin-27 into Myeloid Progenitors to Control Infection in Emergency Myelopoiesis
|
Dimensionality reduction has been applied in various brain areas to study the activity of populations of neurons . To interpret the outputs of dimensionality reduction , it is important to first understand its outputs for brain areas for which the relationship between the stimulus and neural response is well characterized . Here , we applied principal component analysis ( PCA ) to trial-averaged neural responses in macaque primary visual cortex ( V1 ) to study two fundamental , population-level questions . First , we characterized how neural complexity relates to stimulus complexity , where complexity is measured using relative comparisons of dimensionality . Second , we assessed the extent to which responses to different stimuli occupy similar dimensions of the population activity space using a novel statistical method . For comparison , we performed the same dimensionality reduction analyses on the activity of a recently-proposed V1 receptive field model and a deep convolutional neural network . Our results show that the dimensionality of the population response changes systematically with alterations in the properties and complexity of the visual stimulus .
Dimensionality reduction has been applied to neural population activity to study decision making [1 , 2] , motor control [3–5] , olfaction [6] , working memory [7 , 8] , visual attention [9] , audition [10] , rule learning [11] , speech [12] , and more ( for a review , see [13] ) . In many cases , dimensionality reduction is applied in brain areas for which the relationship between neural activity and external variables , such as the sensory stimulus or behavior , is not well characterized . This is indeed the setting in which dimensionality reduction may be most beneficial because it allows one to relate the activity of a neuron to the activity of other recorded neurons , without needing to assume a moment-by-moment relationship with external variables . However , it is also the setting in which the outputs of dimensionality reduction can be the most difficult to interpret . To aid in interpreting the outputs of dimensionality reduction in such settings , it is important to vary the inputs to a brain area and ask whether the outputs of dimensionality reduction change in a sensible way . This is most readily done for a brain area close to the sensory periphery , such as the primary visual cortex ( V1 ) . Here , we apply dimensionality reduction to V1 and ask two fundamental , population-level questions . First , how is neural complexity related to stimulus or task complexity ? Previous studies have used dimensionality reduction to analyze population activity in a reduced space ( e . g . , [1–13] ) . Implicit in these studies is the appropriate dimensionality of the reduced space , which is a measure of neural complexity . It is currently unknown how neural complexity scales with stimulus or task complexity for a given population of neurons [14] . Second , how does a neural circuit flexibly encode ( or “multiplex” ) the representation of the vast number of stimuli encountered in the natural world ? Recent studies suggest that it may be possible to take advantage of the multi-dimensional properties of the population activity space [2 , 7 , 15–17] . In particular , the population activity representing different stimuli might occupy similar dimensions of the population activity space [10] . It is currently unknown how the similarity of the dimensions being occupied by the population activity changes with the similarity of the stimuli . The concept of dimensionality is illustrated in Fig 1 . Consider a high-dimensional space ( termed the population firing rate space ) in which each axis represents the firing rate of a recorded neuron ( Fig 1A ) . The goal of dimensionality reduction is to identify i ) how many dimensions are occupied by the neural population activity , i . e . , the dimensionality of the population activity , and ii ) how these dimensions are oriented within the population firing rate space . In this three-neuron example , the population activity is two-dimensional , where the dimensions are defined by the orthogonal basis patterns ( Fig 1A , basis patterns 1 and 2 ) . Equivalently , we can think of dimensionality reduction in terms of decomposing the population activity into a weighted sum of basis patterns and a mean offset ( Fig 1B ) . A basis pattern describes a characteristic way in which activity of the neurons covaries . Each basis pattern is fixed and is weighted by a time-varying latent variable , which represents the contribution of the basis pattern at each point in time . We can compare the outputs of dimensionality reduction for two different stimuli ( Fig 1B and 1C ) . The population activity for stimulus 1 is two-dimensional because it can be described by two basis patterns ( Fig 1B ) , whereas that for stimulus 2 is three-dimensional ( Fig 1C ) . Thus , the population response to stimulus 2 would be deemed more complex than the population response to stimulus 1 . In addition , we can ask whether the population responses to different stimuli occupy similar dimensions within the population firing rate space . This is assessed by comparing the basis patterns across stimuli . In this example , there is one basis pattern that is shared by both stimuli ( red ) , one basis pattern that is similar between stimuli ( green ) , and one basis pattern ( black ) that is employed only by stimulus 2 . In this work , we characterize how neural complexity varies with stimulus complexity in macaque V1 by applying principal components analysis ( PCA ) to the trial-averaged neural responses to different classes of visual stimuli , including sinusoidal gratings , a natural movie , and white noise . In addition , we develop a new method ( termed the pattern aggregation method ) to measure how the basis patterns extracted from the population responses to each stimulus relate to each other . This method allows one to characterize how the similarity of the dimensions being occupied by the population activity changes with the similarity of the stimuli . A key advantage of studying these questions in V1 is that there are well-established receptive field ( RF ) models . By applying the same dimensionality reduction methods to the outputs of an RF model , we can more deeply understand the relationship between the outputs of dimensionality reduction and known properties of V1 neurons . The results described in this work show that the outputs of dimensionality reduction , when applied to V1 population activity , are sensible , and thus may be fruitfully applied to other brain areas .
We first investigated how changing the complexity of the visual stimulus changes the dimensionality of trial-averaged population responses using drifting sinusoidal gratings . To change the stimulus complexity , we included different numbers of consecutive grating orientations in the analysis ( ranging from 1 to 12 orientations ) . For example , the least complex stimulus included a single orientation , and more complex stimuli included two or five consecutive orientations ( Fig 2A ) . We asked how quickly the dimensionality of the population activity grows as the number of orientations increases . At one extreme , it may be that the population response to each orientation uses an entirely different set of basis patterns ( i . e . , dimensions ) . In this case , the dimensionality for two orientations would be two times the dimensionality for one orientation . At the other extreme , it may be that the population response to each orientation resides in the same set of dimensions . In other words , the population response is formed using the same basis patterns , but linearly combined using different weights for different orientations . In this case , the dimensionality for two orientations would be the same as the dimensionality for one orientation . We first computed the basis patterns of each orientation individually by applying PCA to the trial-averaged population response ( taken in 20 ms bins ) , and identifying the patterns explaining the greatest variance in the population response ( up to a chosen cumulative variance threshold , e . g . , 90% ) . To assess the dimensionality of the population response to multiple orientations , we developed the pattern aggregation method ( see Methods ) , which first aggregates the basis patterns for different orientations as column vectors in a matrix , and then computes the number of linearly independent columns of the aggregated matrix ( i . e . , the effective rank ) . This value is the dimensionality for multiple orientations . Using a 90% variance threshold , we found that the dimensionality for two orientations was 1 . 62 times the dimensionality for one orientation ( Fig 2B , ‘90%’ curve ) , and significantly smaller than what would be expected had the basis patterns been randomly chosen ( Fig 2B , ‘chance 90%’ curve , p < 10−5 ) . In other words , for consecutive orientations separated by 30° , the population responses share about half of their basis patterns . As more orientations were included , the dimensionality of population responses remained lower than expected by chance ( Fig 2B ) , indicating that population responses to oriented gratings separated by angles larger than 30° also tend to use similar basis patterns . Similar trends were found using different variance thresholds ( Fig 2B , ‘70%’ , ‘80%’ curves ) , so we use a 90% threshold in the rest of this work . The pattern aggregation method requires a choice for the rank threshold t to determine how different basis patterns need to be before they define separate dimensions . We repeated the above analysis for different choices of t and a fixed variance threshold of 90% ( Fig 2C ) . We found that the dimensionality trends are similar for rank thresholds t near 0 . 5 , so we use t = 0 . 5 in the rest of this work . Because the absolute dimensionality depends on the variance and rank thresholds , we make no claims about absolute dimensionality . Rather , we focus on relative comparisons of dimensionality for fixed variance and rank thresholds . We observed a change in the rate of increase of dimensionality after six orientations ( Fig 2B , black curve ) . Because consecutive orientations were separated by 30° , the first and seventh orientations were 180° apart and differed only in their drift direction . Thus , the seventh to twelfth orientations were identical to the first to sixth orientations respectively , but drifted in opposite directions . A small proportion of V1 neurons are direction selective [18 , 19] , so the change in slope of the dimensionality curve might be due to the population activity using similar basis patterns for opposite drift directions . To test this possibility , we performed two analyses . First , we assessed the direction selectivity of each neuron ( direction index = 1 − null response / preferred response ) , and found that 16 of 183 neurons had a direction index greater than 0 . 5 , consistent with [19] . If none of the neurons encoded direction selectivity , we would expect the dimensionality curve to be flat beyond 6 orientations in Fig 2B . The increase in dimensionality after 6 orientations is consistent with the finding that at least some neurons show direction selectivity . A potential concern is that the increase in dimensionality beyond 6 orientations arises from the fact that a larger number of patterns are being aggregated for a larger number of orientations . To address this , we performed a control analysis which equalized the number of patterns being aggregated across different numbers of conditions by including patterns from many subsamples of the data . We found that the dimensionalities for 7 or more orientations were significantly greater than the dimensionality for 6 orientations ( p < 0 . 001 ) , following the same trend shown in Fig 2B . Second , we assessed how the dimensionality of the population activity for two orientations varies with the angular offset between the orientations ( Fig 2D ) . This indicates how the similarity of the dimensions being occupied by the population activity changes with the similarity of the stimuli . We found that the dimensionality increases with angular offset up to 90° , indicating that the population activity differs the most for two orientations with 90° offset . Then , the dimensionality decreases as the angular offset increases , reaching a minimum at a 180° offset , where gratings drift in opposite directions . Thus , the population activity uses more similar basis patterns for gratings drifting in opposite directions ( 180° offset ) than to gratings of different orientations ( angular offsets other than 180° ) . The dimensionality for 180° offset was higher than that for 0° offset ( p < 10−5 ) , indicating that the population activity does not use identical basis patterns for opposite drift directions . Because these dimensionalities were computed differently ( the pattern aggregation method for 180° offset and a variance threshold for 0° offset ) , we aggregated an equal number of patterns ( identified over many subsampled runs ) for 0° offset as that for 180° offset , and still found a higher dimensionality for 180° offset than for 0° offset ( p < 10−5 ) . This result , combined with the change in slope of the dimensionality curve ( Fig 2B ) , indicates that the population activity tends to use similar ( but not identical ) basis patterns for opposite drift directions . Taken together , the analyses in Fig 2B and 2D characterize how the outputs of dimensionality reduction vary with the sensory input ( in this case , drifting sinusoidal gratings ) to V1 . We also visualized the basis patterns describing the largest percentage of variance for three different orientations ( Fig 2E ) . These basis patterns extracted from the trial-averaged population activity are akin to the hypothetical basis patterns shown in Fig 1 ( red , green , black ) . For a given basis pattern , the absolute height of each bar indicates the degree to which each neuron contributes to that basis pattern . The following are two salient properties of the identified basis patterns . First , most of the neurons in the recorded population contribute to each basis pattern to some extent . Thus , the basis patterns capture changes in firing rates that are shared broadly across the population , rather than reflecting the activity of only a small number of neurons . Second , the basis patterns capture both positive and negative signal correlations between neurons , where the signal is the phase of the grating at different time points . A basis pattern describes positive signal correlation between a pair of neurons if the neurons have coefficients of the same sign . Conversely , a basis pattern describes negative signal correlation for coefficients of opposite sign . We can relate these basis patterns to the results shown in Fig 2B–2D by asking how similar are the linear combinations of each set of basis patterns across different stimulus orientations . This is difficult to assess by eye , and so we rely on the quantifications shown in Fig 2B–2D to determine how similar are the basis patterns across stimulus orientations . We next sought to determine how the dimensionality of the trial-averaged population activity varies with the complexity of the visual stimulus for a wider range of stimuli . We presented three movie stimuli ( Fig 3A ) : a sequence of sinusoidal gratings ( ‘gratings movie’ ) , contiguous natural scenes ( ‘natural movie’ ) , and white noise ( ‘noise movie’ ) . In contrast to Fig 2A where the order of stimulus complexity is clear ( i . e . , a larger number of orientations is more complex ) , here we needed first to assess the relative complexity of the three movie stimuli . By applying PCA to the pixel intensities , we found that 40 dimensions could explain nearly 100% of the variance of the pixel intensities for the gratings movie ( Fig 3B ) . For the natural movie , the top few dimensions explained a large percentage of the variance due to global luminance changes caused by zooming and panning the camera , and a large number of additional dimensions were needed to explain the remaining variance . For the noise movie , each dimension explained only a small percentage of the total variance . We summarized these cumulative percent variance curves by finding the number of dimensions ( gratings movie: 24 , natural movie: 64 , noise movie: 459 ) needed to explain 90% of the variance ( Fig 3B , dashed line ) . Based on these values , the gratings movie was least complex , followed by the natural movie , then the noise movie . Similar results were obtained when first transforming the pixel intensities using a V1 receptive field model , then applying PCA ( see “Comparing to a V1 receptive field model” ) . We further asked how much the pixel intensities varied for each movie stimulus , and found that the noise movie had a smaller variance than the other two movie stimuli ( Fig 3B , inset ) . Together , this indicates that the distribution of pixel intensities for the gratings movie and natural movie is akin to an elongated ellipsoid ( low dimensionality , high variance ) , whereas that for the noise movie is akin to a small ball ( high dimensionality , low variance ) . Having measured the relative complexity of the movie stimuli , we then asked how the dimensionality of the population responses to the movie stimuli varies with stimulus complexity . We found that for a 90% variance threshold , the dimensionality of the trial-averaged population responses ( 20 ms bins ) was ordered in the same way as the stimulus complexity ( Fig 3C ) ; namely , the population responses to the gratings movie had the lowest dimensionality , followed by the natural movie , then the noise movie . This ordering did not simply follow from the ordering of the mean population firing rates for the different movies ( monkey 1: 4 . 2 , 6 . 4 , 5 . 4 spikes/sec , monkey 2: 6 . 6 , 8 . 2 , 6 . 7 spikes/sec for gratings , natural , and noise movies , respectively ) , and was consistent for a wide range of neuron counts for both monkeys ( S1 Fig ) . We also assessed how much the firing rates varied in response to each movie stimulus—that is , we measured how much the mean firing rate ( averaged across experimental trials ) varies over time . As with pixel intensities , we found that the population response to the noise movie had the smallest variance , followed by the gratings movie , then the natural movie ( Fig 3C , inset ) . Overall , the dimensionality and variance ordering in the visual stimuli and the population responses were similar , indicating that the population activity in V1 retains the complexity of the ordering of the visual stimuli themselves . Having compared the dimensionality of the population activity across stimuli , we next asked how the basis patterns of the population activity ( corresponding to the dimensions being occupied by the population activity ) compare across stimuli . Previous studies have found that the ability of a RF model to predict a neuron’s response can depend on the stimulus class on which the model was trained [20–22] , suggesting that the population activity might use somewhat distinct basis patterns for different stimulus classes . On the other hand , if basis patterns are influenced by the shared underlying network structure [4 , 10] , then we would expect them to be shared across responses to different stimuli . We first asked whether there are qualitative differences in the coefficients of the basis patterns for population responses to the stimulus movies ( Fig 4A ) . As in Fig 2E , we found that most of the basis patterns represented activity across a large number of neurons and described both positive and negative signal correlations . However , there were two notable exceptions . First , the basis pattern describing the most variance for the population response to the gratings and noise movies involved primarily two neurons ( Fig 4A , right and left panels , pattern 1 , neuron indices 26 and 27 ) . For these movies , the two neurons had the highest firing rate modulation ( maximum—minimum firing rate ) across the recorded population . The weights for the gratings movie appeared to be sparser than those for individual gratings ( Fig 2E ) , likely because the gratings movie contained different orientations that strongly co-modulated these similarly-tuned neurons , whereas individual gratings with a single orientation co-modulated many neurons together with phase . Second , the basis pattern describing the most variance for the population response to the natural movie had coefficients of the same sign ( Fig 4A , middle panel , pattern 1 ) . This is due to the entire population increasing or decreasing its firing rates together in response to global luminance changes prevalent in natural movies . Other than the similarity of the top basis pattern for the gratings and noise movies , it was difficult to determine by eye whether the basis patterns were being shared across stimuli . Thus , we used the pattern aggregation method to quantitatively assess the similarity of the identified basis patterns . As a baseline , we assessed the extent to which the visual stimuli themselves reside in the same dimensions in pixel space . We compared the dimensionalities of the individual movies ( Fig 4B , teal dots , consistent with Fig 3B ) to those of combinations of movies ( Fig 4B , orange and purple dots ) . If the stimuli reside in overlapping dimensions , then the resulting dimensionality would be the maximum of the dimensionalities for the individual movies . However , if the stimuli reside in completely non-overlapping ( i . e . , orthogonal ) dimensions , then the resulting dimensionality would be the sum of the dimensionalities for the individual movies . To measure the extent of overlap , we computed a similarity index s , for which s > 0 indicates that the patterns are more similar ( i . e . , more overlapping ) than expected by chance and s < 0 indicates that the patterns are less similar ( i . e . , closer to orthogonal ) than expected by chance ( see Methods ) . There are two main observations . First , the dimensions occupied by the gratings movie overlap with those for the natural movie . To see this , note that the dimensionality for the gratings and natural movies together ( 80 dimensions ) was less than the sum of the individual dimensionalities for the two movies ( Fig 4B , gray , 88 dimensions ) . To ensure that the overlap in patterns was meaningful , we confirmed that the aggregated dimensionality of 80 was less than the dimensionality ( 88 dimensions ) that would be expected by combining randomly-chosen dimensions ( s = 0 . 33 , p < 10−5 ) . Second , the dimensions occupied by the noise movie include many of the dimensions for the other two stimuli . This is indicated by the fact that the dimensionality corresponding to any combination of stimuli that included the noise movie ( gratings + noise: 472 , natural + noise: 495 , and gratings + natural + noise: 500 dimensions ) was less than the dimensionality that would be expected by chance ( s > 0 . 4 , p < 10−5 for all cases ) . Note that , in all cases , the chance dimensionality was near the maximum dimensionality ( indicating orthogonality between the two sets of randomly-chosen patterns ) because the dimensionality of the pixel space ( 1 , 600 dimensions ) was much larger than the dimensionalities of the individual movies . We used the same approach to analyze the population responses as we did the visual stimuli . We compared the dimensionality of the population responses to individual movies ( Fig 4C , teal dots , consistent with Fig 3C ) to that of population responses to combinations of movies ( Fig 4C , orange and purple dots ) . We found that the relationship of the basis patterns employed by the population activity across stimuli ( Fig 4C ) was similar to the relationship between the stimuli themselves ( Fig 4B ) . First , the dimensions occupied by the population responses to the gratings movie are overlapping with those for the natural movie . This is indicated by the fact that the aggregated dimensionality for the population responses to the gratings and natural movies ( monkey 1: 25 , monkey 2: 38 ) was less than the dimensionality if the patterns were orthogonal ( top of the black brackets ) and the dimensionality expected by chance ( s > 0 . 39 , p < 10−5 ) . Second , the dimensions occupied by the population response to the noise movie include most of the dimensions for the other two stimuli . This is indicated by the fact that the dimensionality corresponding to any combination of stimuli that included the noise movie was less than the dimensionality expected by chance ( s > 0 . 32 , p < 10−5 in all cases ) , and close to entirely overlapping . The chance dimensionality in Fig 4C is computed by assuming that any population activity pattern within the N-dimensional population firing rate space can be achieved ( where N = 61 for monkey 1 and N = 81 for monkey 2 ) . Previous studies indicate that the population activity may only be able to occupy a subset of dimensions due to underlying network constraints [4 , 10] . Although we had no way of identifying exactly how many dimensions could have been occupied by the population of recorded neurons , we computed a lower bound by determining M , the largest value of dimensionality observed in response to any combination of stimuli ( Fig 4C , M = 39 for monkey 1 and M = 52 for monkey 2 ) . If the chance level is computed instead by drawing random patterns from this M-dimensional space , we still find that the population activity tends to occupy more similar dimensions than expected by chance for all pairs of movies ( p < 0 . 05 ) . This is a conservative assessment because larger values of M would only make the comparison more statistically significant . We note that even if the population responses to different stimuli occupy similar dimensions , this does not imply that the responses occupy the same regions of the subspace defined by those dimensions . In other words , the activity may covary along the same dimensions but be centered at different locations in the population activity space . We found that the population responses to the three movies indeed were centered in different locations of the population activity space ( S2 Fig ) . Our analysis of the similarity of basis patterns for the population activity is summarized by the schematic in Fig 4D , where the size of each ellipse represents the dimensionality ( i . e . , number of basis patterns ) of the population response to the corresponding stimulus and the overlap between ellipses represents the extent to which the population responses share basis patterns . We found that the basis patterns for the gratings movie were overlapping with those of the natural movie , and that the basis patterns for the noise movie largely contain the basis patterns for the other two stimuli . However , there were a small number of basis patterns that were unique to each stimulus , shown by the small areas of non-overlap among the ellipses . Overall , this suggests that a neural circuit is capable of expressing a limited repertoire of basis patterns , and that a subset of those basis patterns is employed for any given stimulus . In the preceding sections , the analyses of the movie stimuli and the corresponding neural activity used the entire 30-second movie ( i . e . , 750 time points ) together . Here , we consider time-resolved measurements of dimensionality using one second windows , each comprising 25 time points . This allows us to assess how dimensionality changes over time , and compare the basis patterns employed by the trial-averaged population activity during different parts of the 30-second movies . For the visual stimuli ( Fig 5A , left panel ) and the population responses ( Fig 5A , center and right panels ) , we found that the dimensionality corresponding to the noise movie was higher than the dimensionality corresponding to the gratings and natural movies in each one-second window , consistent with the results of analyzing each 30-second movie in its entirety ( Fig 3 ) . However , the dimensionality of the natural movie was not greater than that of the gratings movie ( cf . colored triangles in Fig 5A , which indicate average dimensionality over time ) , in contrast to Fig 3 . We hypothesized that this was due to temporal correlations in the natural movie , in which frames within a one-second window tend to be self-similar . In contrast , for the gratings and noise movies , there are at least three different grating orientations and 25 different frames of white noise within each one-second window . To reconcile the results for short and long time windows , we performed three analyses . First , we tested the hypothesis that temporal correlations in the natural movie result in lower dimensionalities ( relative to the other movies ) for short time windows . To break the temporal correlations , we shuffled the time points ( 20 ms resolution ) across each 30 second period and performed the same analysis as in Fig 5A . We found that , in a one-second window , the mean dimensionality corresponding to the natural movie was higher than that corresponding to the gratings movie for the shuffled data . This was true for the visual stimuli ( p < 0 . 01 ) and for the population responses ( monkey 1: p < 10−3 , monkey 2: p < 0 . 05 ) . This indicates that the range of basis patterns expressed by the population responses to the natural movie within a short time window is limited by the temporal correlations in the natural movie itself . Second , we asked how the dimensionality grows when increasing the window size progressively from one second to 30 seconds , where each window starts at the beginning of the movie ( Fig 5B ) . If the dimensionality increases with window size , this would indicate that new patterns ( of pixels or of population activity ) are being used throughout the duration of the movie . However , if the dimensionality plateaus , then the patterns are being reused and no new patterns are being expressed . The leftmost points on these curves ( one-second window ) correspond to the leftmost points on the corresponding curves in Fig 5A . The rightmost points on these curves ( 30-second window ) correspond to the dimensionalities shown in Fig 3 . Although the dimensionality corresponding to the natural movie ( green ) is lower than that corresponding to the gratings movie ( blue ) for short time windows , the dimensionality corresponding to the natural movie grows more quickly and surpasses the dimensionality corresponding to the gratings movie for longer time windows . We found this to be true for both the visual stimuli ( Fig 5B , left panel ) and the population responses ( Fig 5B , center and right panels ) . This indicates that the patterns for the natural movie tend to be self-similar within a short time window , and new patterns are expressed over longer time windows . In contrast , the dimensionality corresponding to the gratings movie does not grow as quickly with window size . This is because once a few different grating orientations are included , patterns corresponding to additional grating orientations can be represented approximately as linear combinations of patterns corresponding to existing grating orientations ( both for the visual stimuli and the population responses ) , and therefore do not increase the dimensionality appreciably . Third , we directly measured the similarity of the patterns between different one-second windows using the pattern aggregation method . Fig 5C shows the matrix of similarity indices for the population responses to each of the three movies ( monkey 1 ) . By averaging the values of the off-diagonal similarity indices across the matrix , we found that the patterns corresponding to the gratings movie tended to be more similar across time than those for the natural and noise movies ( Fig 5D ) . This was true for the visual stimuli ( left panel ) , as well as for the population responses ( middle and right panels ) . This result indicates that new patterns tend to be expressed ( for both the visual stimuli and the population responses ) as the movies play out over time more for the natural and noise movies than for the gratings movie . This result also supports the finding in Fig 5B that the dimensionality corresponding to the natural movie grows more quickly than that corresponding to the gratings movie . Taken together , these results indicate that the noise movie drives a large number of basis patterns in the population activity in a short time window , relative to the gratings and natural movies . As the movies play out over time , the noise and natural movies tend to drive new basis patterns , whereas the grating movie tends to recruit the same patterns . We also asked whether the second-by-second fluctuations of the dimensionality of the visual stimulus during the 30-second movie ( Fig 5A , left panel ) were related to that of the population responses ( Fig 5A , middle and right panels ) and found weak correlations ( mean across movies ρ = 0 . 20 ) . Although the dimensionality fluctuations were larger for the natural movie than the other two movies ( Fig 5A , left panel ) , these fluctuations were less salient in the population activity ( Fig 5A , middle and right panels ) . These two observations underscore that , although there are similarities in the statistical properties of the visual stimuli and population responses , there are also differences that remain to be understood . One of the key advantages of performing this study in V1 is that much is known about the stimulus-response relationship of individual neurons , as described by the many RF models that have been proposed in the literature [23] . Although the RF models do not capture every aspect of V1 neuronal activity [20–22 , 24] , we can apply the same dimensionality reduction methods to the activity generated by an RF model to help interpret the outputs of dimensionality reduction . We consider it a strength that , in many cases ( described below ) , the outputs of dimensionality reduction applied to population activity show the same trends as when applied to activity from an RF model . This similarity indicates that single-neuron properties are reflected in population metrics , such as dimensionality . In cases where there are discrepancies between the outputs of dimensionality reduction for population activity and RF models , our results can provide guidance necessary to improve RF models . Although a complete study of the many V1 RF models available is beyond the scope of this work , here we focus on one recently-proposed model [25] . The model has four components: a Gabor filter , whose output is half-rectified; an untuned suppressive filter , whose output is also half-rectified; a normalization signal; and an exponentiating output nonlinearity . These components are shared with many other models of V1 [23] , and thus make it well-suited for studying how dimensionality changes as the input image is transformed by each component . The parameter values for 100 model neurons were drawn from parameter distributions reported in [25] , since we did not present the stimuli appropriate for fitting the model parameters to data . For this reason , we only compare trends between the dimensionality of the outputs from the model and that of population activity . We first assessed the dimensionality of each component’s responses to the same individual gratings as presented to the monkeys . We observed similar trends between the population activity ( Fig 2B and 2D ) and the activity of the RF model ( Fig 6B and 6C ) . As expected , the model captures key aspects of the population response to gratings , including direction selectivity . However , the last component of the model ( “pointwise nonlinearity” ) showed substantially smaller dimensionalities than the other components ( Fig 6B and 6C , rightmost panels ) . This decrease in dimensionality is due to the exponential function exaggerating the anisotropy of the distribution of the model activity . For example , if the distribution of the responses across images resembles an ellipsoid in the 100-dimensional firing rate space , the exponential function would expand the variances of the major axes considerably more than the variances of the minor axes . The major axes would explain a greater proportion of the overall variance , resulting in fewer dimensions . Because we selected the values of the exponents independently from other model parameters , the discrepancy in dimensionality for the last component might not be present in the original model in [25] , whose parameters were fit together . Still , our results indicate dimensionality can be sensitive to some nonlinear transformations . Next , we assessed the dimensionality of each component’s responses to the same movie stimuli as presented to the monkeys . The ordering of dimensionality for each component of the model ( Fig 6D ) followed that of the population activity ( Fig 3C ) . As for the individual gratings , the pointwise nonlinearity substantially reduced the dimensionality of the model activity . A discrepancy between the model and the population activity was the ordering of variance: each component of the model exhibited greater variance for the gratings movie than for the natural movie ( Fig 6D , insets ) , but this was not the case for population activity ( Fig 3C , insets ) . This discrepancy stems from the oriented Gabor filters in the first component of the model . Because the temporal frequency of these filters was matched to that of the drifting sinuosoid gratings , the filters were modulated strongly by the gratings movie . One advantage of working with an RF model is that we can assess the dimensionality of the responses to images that were not shown to the monkeys . With the caveat that the RF model might not capture aspects of the responses of real neurons , we assessed how the dimensionality of each component’s outputs varies as we parametrically alter the visual stimuli . We first considered varying the contrast of the natural movie images . Initially , one might think that contrast would have no effect on dimensionality because each dimension in pixel space is scaled equally . However , this is not the case for two reasons . First , before scaling , each image was re-centered by subtracting out the image’s mean pixel intensity—not the mean pixel intensity across all images . Thus , changing the contrast is not a simple scaling across all images . The second reason is that there are nonlinearities in the model ( e . g . , the linear rectifying functions ) , which can “warp” the scatter of data points ( where each point corresponds to an image ) , thereby changing the dimensionality . When we varied the contrast of the natural movie images , we found that dimensionality decreased with decreasing contrast ( different colors in Fig 7A ) . This decrease in dimensionality occurs because , as contrast decreases , the mean luminance dominates each image . At a contrast of 0% , the images lie along a line in pixel space ( i . e . , the [1 , 1 , … , 1] direction ) . Next , we randomized the phases of each natural image by adding a random offset to each phase value . The offsets were drawn from uniform distributions of varying extent . We considered small offsets drawn from the range [−45° , 45°] to large offsets drawn from the range [−180° , 180°] . When the phases were completely random ( i . e . , a range of [−180° , 180°] ) , the statistics of the images were akin to pink noise [26] . We observed that as the extent of the phase randomization increased , the dimensionality decreased ( Fig 7B ) . Although randomizing the phases removed higher-order spatial correlations ( e . g . , edges and textures ) , it did not remove the spatial correlations brought about by the low frequencies of the power spectrum , which are strongly represented in natural images [27] . Only a small number of dimensions are needed to capture these low-frequency spatial correlations , because most of the pixels inside the region covered by the RFs of the model’s filters have pixel intensities that co-vary strongly . Finally , to remove these low-frequency spatial correlations , we gradually transformed each phase-randomized image ( i . e . , pink noise image ) to a white noise image by raising the power spectrum of the pink noise images to a fractional exponent . As the power spectrum of the pink noise images became more similar to white noise , the dimensionality increased ( Fig 7C , black to red ) . This increase is not because the model neurons were more responsive to white noise images than pink noise images . Instead , the outputs of the model expressed more activity patterns for the white noise images due to the uncorrelated pixel intensities . This is consistent with the dimensionality trends of the population activity , where we observed that the population response to white noise had the highest dimensionality of the three movies but the lowest amount of variance ( Fig 3C ) . Based on these results , the model predicts that lowering contrast and randomizing the phases of a set of natural images will decrease the dimensionality of the population response . We also observed that for all manipulations of the visual stimuli , the outputs of the last component of the model ( pointwise nonlinearity ) showed substantially lower dimensionality than outputs at other model components , consistent with the results in Fig 6 . This suggests that certain nonlinearities ( e . g . , pointwise exponentiation ) affect dimensionality more than others ( e . g . , divisive normalization ) . To build intuition about how the dimensionality of population activity might change at different stages of visual processing , we examined a deep convolutional neural network that was previously trained with over 1 million natural images and showed high accuracy on a well-known image recognition challenge [28] . The deep network takes an image as input , processes the image through layers of filtering and nonlinear operations , including convolution , pooling , and normalization ( Fig 8A ) . The earlier layers capture low-level image statistics , such as oriented edges , while the deeper layers capture high-level image statistics , such as features that distinguish a car from a building [28] . This hierarchical processing resembles the processing found in the visual system , and indeed , the progressive layers of deep networks appear to mimic the progressive processing stages of the ventral stream [29] . We used this deep network to assess how dimensionality changes from one layer to the next , providing a prediction of how dimensionality might change in different visual areas along the ventral stream . We input the same movie stimuli that we presented to the monkeys into the deep network , and examined the dimensionality of the filter outputs in different layers of the deep network ( Fig 8B ) . For each layer , we analyzed the 100 filters that had RFs closest to the center of the image . We found that the earliest layer ( Fig 8B , layer 1 ) showed dimensionality and response variance trends that matched the V1 population activity ( Fig 3C ) . In progressively deeper layers ( Fig 8B , going from layers 1 to 8 ) , the responses to the gratings and noise movies decreased in dimensionality and relative variance , whereas the responses to the natural movie increased in dimensionality and relative variance . These findings are consistent with our understanding of the ventral stream: in progressive stages of visual processing , neurons become more sensitive to features in natural images and less sensitive to artificial images [30] . These results thus provide a prediction of how the dimensionality and variance of trial-averaged population responses to natural and artificial images should change along the ventral stream , which can be tested in future experiments [17] .
To aid in understanding the outputs of dimensionality reduction , we chose to study a brain area close to the sensory periphery ( V1 ) . This allowed us to vary the sensory inputs and ask whether the outputs of dimensionality reduction change in a sensible way . By applying PCA to trial-averaged population responses to different classes of visual stimuli , including sinusoidal gratings , a natural movie , and white noise , we found that the dimensionality of the population responses grows with the stimulus complexity . In addition , we assessed whether the population responses to different stimuli occupy similar dimensions of the population firing rate space using a novel statistical method ( the pattern aggregation method ) . We found that the population responses to stimuli as different as gratings and natural movies tended to occupy similar dimensions . For comparison , we applied the same analyses to the activity of a recently-proposed V1 receptive field model and a deep convolutional neural network , both of which showed trends similar to the real data . We further used these models to predict the dimensionality trends of the population responses to visual stimuli not shown to the monkeys , as well as the dimensionality trends of population responses in brain areas other than V1 . Many previous studies have compared visual cortical responses to natural and artificial stimuli on a single-neuron level ( e . g . , [20–22 , 31 , 32] ) . Their predominant approach was to define a parameterized RF model to relate a neuron’s activity to the visual stimulus . These studies found that , although RF models derived from natural stimuli share properties with those derived from artificial stimuli [20 , 31] , there can be important differences [20–22] . Here , rather than relating each neuron’s activity to the stimulus , we relate the activity of the recorded neurons to each other . We can then ask how this relationship ( i . e . , the covariation of trial-averaged activity among neurons ) changes for different classes of visual stimuli . This approach has been adopted for pairs of neurons ( i . e . , signal correlation ) [33] , and we extend this work to characterize the signal correlation among all pairs of neurons at once . We found that the gratings , natural , and noise stimuli elicit many common basis patterns , consistent with previous studies showing similarities between RFs measured with gratings and natural stimuli [20] and those measured with natural and noise stimuli [31] . Finally , our finding of some unique basis patterns for each stimulus class suggests that estimates of RFs will best capture the responses to the same type of stimulus used in estimating the RFs , as reported in previous studies [21 , 22] . The quality of most RF models has been evaluated based on their ability to predict the activity of individual neurons ( e . g . , [20–23 , 25 , 31] ) . Given that there can be , in some cases , a substantial difference between the predictions of RF models and the recorded neural activity [21 , 22 , 31] , especially for natural scenes , we need to quantify how they are different in an effort to improve the RF models . This is often quantified by computing the percent variance of the recorded activity explained by the model for each neuron individually [21–23] . Our work provides a complementary way to compare RF models and recorded activity by examining the entire population together . We can compare the many V1 models that have been proposed by assessing which ones best reproduce the relative dimensionalities across stimuli and the similarity of basis patterns observed in recorded activity . The model that we tested does reproduce the dimensionality trend of the population activity across stimuli , but does not reproduce the response variance trend ( Fig 6D ) . We speculate that a different spatiotemporal filter in the first component of the model can help to increase the response variance to natural images [21] , thereby better matching the response variance trend of the model to that of the recorded activity . Different basis patterns may be used by the population activity during different task epochs , suggesting that certain basis patterns drive downstream areas more effectively than others [16 , 34] . Thus , the identification of which basis patterns are used may be critical for understanding how different brain areas interact on a population level [35] . Furthermore , the activity patterns across a neural population have been used to study normalization [36] , decision making [2 , 37] , learning [4] , and motor planning [5] . In the present work , we have developed a statistical framework ( the pattern aggregation method ) to measure the similarity of basis patterns across any number of stimulus conditions or time points . We validated this framework using recordings in V1 , and the framework can be applied broadly to other brain areas . The measurement of dimensionality depends on many factors , including the choice of dimensionality reduction method , the number of neurons ( cf . S1 Fig ) , and the number of data points . In principle , one should use a nonlinear dimensionality reduction method ( e . g . , [38–40] ) because the underlying manifold of the population activity is likely to be nonlinear . For example , divisive normalization nonlinearly maps the tuning curves of a population of neurons onto a high-dimensional sphere [41] , and a nonlinear dimensionality reduction method may be able to extract the lower-dimensional embedding . However , most studies using dimensionality reduction in systems neuroscience have focused on linear methods [1–4 , 6 , 11 , 12 , 16 , 42] . The reasons are that 1 ) most nonlinear methods rely on a dense sampling of the population activity space , in contrast to experimental data which tend to sparsely sample the space , and 2 ) it is usually difficult to assess the contribution of each neuron to a low-dimensional space identified by nonlinear methods . For the latter reason , we would not be able to compare how similar are the patterns for different stimuli , as we do in this study . Despite these caveats , we applied a nonlinear method , fractal dimensionality [17 , 38] , to the three movies and their population responses ( S3 Fig ) . The ordering of fractal dimensionality across stimuli was consistent with that of PCA dimensionality ( Fig 3 ) . Together with the results showing how the dimensionality ordering of the population activity depends on stimulus complexity ( Figs 2B and 3C ) and neuron count ( S1 Fig ) , this finding indicates that a linear method can provide useful insights , even if the underlying manifold is indeed nonlinear . There are several other factors that can affect the dimensionality of population responses . Dimensionality can depend on the properties of the particular neurons being sampled . In V1 , these properties include the size and scatter of the receptive fields , as well as their preferred phases , orientations , and spatial frequencies . Another factor that can affect dimensionality in V1 is the size of the visual stimulus . We presented large visual stimuli that extended outside of the classical RF of most neurons . Previous studies have shown that stimulation outside of the classical RF tends to increase the sparseness of V1 responses [43–46] , which may affect the dimensionality of the population response . Sparseness leads to independence in the responses between neurons [43 , 44 , 47] , and may lead to increased discriminability of the population activity [48] . Independence implies that each basis pattern only captures modulations of a single neuron ( i . e . , only one element of each pattern is non-zero ) , and the dimensionality ( as assessed by PCA ) depends on the relative variances captured by the basis patterns ( in this case , the relative variances of the neurons ) . To our knowledge , there is no general relationship between sparseness and dimensionality . For all these reasons , it is not possible to make absolute statements about the dimensionality of V1 . Instead , we made relative comparisons where all of the factors affecting dimensionality are fixed , except for the stimulus content . Although we believe that the results shown here are representative of a wide range of gratings , natural , or noise stimuli , they should be interpreted in the context of the particular visual stimuli used . For example , the dimensionality of the gratings movie and its population response could be increased by including more than one spatial and temporal frequency . Similarly , the dimensionality of the natural movie and its population response likely depends on the particular movie clip shown . If the scenes in the movie change more quickly ( or slowly ) over time , then we would expect the dimensionality over a 30-second time window to be larger ( or smaller ) . For the noise movie , our results in Fig 5B indicate that showing more instances of white noise is not likely to further increase the dimensionality of the population response . However , changing the statistics of the noise in the pixels could change the dimensionality of the population response ( Fig 7C ) . At a population level , several studies have compared visual cortical activity evoked by natural and artificial stimuli to spontaneous activity [49–51] . These studies focused on the raw neural activity , which includes both the trial-averaged component ( i . e . , the PSTHs ) and trial-to-trial variability . Here , we focused on the trial-averaged component . Because trial-to-trial variability can be substantial relative to the trial-averaged component [52 , 53] , it is difficult to directly compare results of these previous studies to those reported here . The current study can be extended to study the population structure of trial-to-trial variability using a dimensionality reduction method such as factor analysis rather than PCA [13] in tandem with the pattern aggregation method . For the stimuli that we tested and the recordings we made , we found that the dimensionality trends were consistent between the visual stimuli and the population responses . However , this need not be the case for other visual stimuli and other brain areas . In fact , a dimensionality trend that is inconsistent between the stimuli and responses may yield important insight into how the stimuli are encoded by the neurons under study . Part of resolving this potential discrepancy may relate to the way in which stimulus complexity is measured . PCA dimensionality captures only a particular aspect of the stimulus , namely the anisotropy of the distribution of pixel intensities in an Euclidian space . Other aspects of the stimuli may influence the neural responses more strongly , and alternate measures of stimulus complexity can be used , for example fractal dimensionality ( S3 Fig ) [17 , 38] or a method based on image features extracted by a deep neural network ( Fig 8 ) . Future studies employing additional stimuli and brain areas can elucidate whether dimensionality trends remain consistent between sensory stimuli and population responses . Our work lays a solid foundation to assess the dimensionality and similarity of basis patterns of neural population activity . Because V1 is a well-studied brain area and is close to the sensory input , our results can be compared with expectations based on our intuition and well-established RF models . Moving forward , these methods can be applied broadly to other brain areas and behavioral tasks to examine how the complexity of the population response changes due to conditions such as attentional state , learning , and contextual modulation .
Details of the neural recordings have been described previously [54 , 55] . Briefly , we recorded from primary visual cortex ( V1 ) of anesthetized , paralyzed macaque male monkeys . Anesthesia was administered throughout the experiment with a continuous intravenous infusion of sufentanil citrate ( 6–18 μg/kg/hr ) . Eye movements were minimized with a continuous intravenous infusion of vecuronium bromide ( 100–150 μg/kg/hr ) . Experiments typically lasted 5–7 days . All experimental procedures followed guidelines approved by the Institutional Animal Care and Use Committee of the Albert Einstein College of Medicine at Yeshiva University , and were in full compliance with the guidelines set forth in the US Public Health Service Guide for the Care and Use of Laboratory Animals . Neural activity was recorded using 96-channel multi-electrode arrays ( Blackrock Microsystems , Salt Lake City , Utah ) , which covered 12 . 96 mm2 and had an electrode length of 1 mm . The electrodes were inserted to a nominal depth of 0 . 6 mm to confine recordings mostly to layers 2–3 . Recordings were performed in parafoveal V1 , with RFs within 5 degrees of the fovea . Voltage waveform segments that passed a separately-chosen threshold for each channel were later spike-sorted offline . For comparisons of population responses to different orientation gratings , we spike sorted responses together across all orientations , thereby obtaining a common set of units . Similarly , we spike sorted responses together across all three movies . We included sorted units for which the voltage waveform had a signal-to-noise ratio ( SNR ) greater than 1 . 5 , where SNR is defined as the ratio of the average waveform amplitude to the standard deviation of the waveform noise [56] . This SNR threshold yielded both single-unit and multi-unit activity ( see [57] for comparison of these signals ) with a median SNR near 2 . 5 for all datasets . We analyzed only neurons with mean firing rates greater than 1 spike per second . We used two sets of visual stimuli . The first set ( termed the individual gratings set ) consisted of individual presentations of drifting sinusoidal gratings with different orientations . The second set ( termed the movies set ) consisted of different classes of visual stimuli , including a sequence of drifting sinusoidal gratings with different orientations , a contiguous sequence of natural scenes , and white noise . All stimuli were presented on a CRT monitor with a frame rate of 100 or 120 Hz , and had mean luminance of approximately 40 cd/m2 . We used a look-up table for all stimuli to correct for the nonlinearity between input voltage and output luminance in the monitor . Error bars for the dimensionality of the population activity were computed by subsampling from all time points . We chose subsampling over bootstrapping , since bootstrapping led to biased estimates due to small sample size relative to the number of neurons . For population responses to the stimulus movies , we randomly subsampled 750 of the 1 , 500 time points , computed the dimensionality of the subsampled points , and repeated this 100 times ( Fig 4C ) . Similarly , for the dimensionality analysis in 1 second windows , we randomly sampled 25 of the 50 time points 100 times ( Fig 5A ) . We did not compute error bars for the visual stimuli because subsampling was not possible ( only 750 available time points ) , and bootstrapping was not possible due to the small number of time points relative to the dimensionality ( 1 , 600 ) of the pixel space . All p-values were computed from 105 runs of a random permutation test . We considered a recently-proposed RF model of V1 neurons [25] that includes four components—oriented Gabor filtering , subtraction of untuned suppressive filtering , divisive normalization , and pointwise nonlinearity—common to many RF models of V1 neurons [23] . We input the individual gratings and movie stimuli into the model , and asked how the dimensionality ordering after each model component compares to that of the population activity . The parameter values for 100 model neurons were drawn from the distributions reported in [25] rather than fit to data because we did not present the mixed gratings stimuli necessary for fitting the parameters . For this reason , we compare dimensionality trends between the model and population activity , rather than their absolute values . To study how dimensionality of the model outputs changes as we parametrically alter the visual stimuli , we performed three analyses . First , we varied the contrast of the images of the natural movie in 5 different increments: 100% , 75% , 50% , 25% , 5% . For each image , we subtracted the mean luminance of that image , scaled the result by one of the percentages above , and added back the mean luminance . Second , we transformed the images of the natural movie to pink noise by randomizing the phase of the 2-d Fourier transform for each image . To avoid removing local contrast of the natural images [26] , we added an offset to each phase , where the offset was randomly drawn from a uniform distribution over the range [−α , α] . We considered 5 values for α: 0° ( i . e . , no change of the natural image ) , 45° , 90° , 135° , 180° ( i . e . , pink noise ) . Third , we transformed the pink noise images ( which retained the power spectrum of the natural images ) to white noise by raising the power spectrum of the pink noise to different fractional exponents . The intuition behind this is that the power spectrum of natural images falls off as 1/f2 , where f is frequency [27] . Raising the power spectrum to a fractional exponent , for example 1/2 , transforms the 1/f2 fall-off to a 1/f fall-off . To ensure the average magnitude of the power spectrum was similar for each transformed image , we normalized the power spectrum of each image by dividing by its sum of magnitudes λ . We then raised the normalized power spectrum to 5 different exponents: 1 ( i . e . , pink noise ) , 3/4 , 1/2 , 1/4 , 0 ( i . e . , white noise ) . Finally , we scaled the resulting power spectrum by λ . To assess how the ordering of dimensionality might change at different stages of visual processing , we studied a deep convolutional neural network ( CNN ) . We used an instantiation of a CNN called GoogLeNet [28] , trained with the Princeton Vision and Robotics Toolkit [60] , and available in Matlab with MatConvNet [61] . The CNN had different processing units , including convolution , pooling , concatenation , and normalization/softmax . Each unit comprised many filters ( from 103 to 106 ) that performed the same operation ( e . g . , convolution ) but on different spatial regions of the input . Each layer comprised a group of units ( shown in Fig 8A ) . We assessed dimensionality of the outputs of eight consecutive layers of the deep network . For the first layer , we analyzed the outputs of 100 filters of the second normalization unit . For layers 2 to 8 , we analyzed the outputs of 100 filters of the concatenation units . For each analyzed unit , we chose the 100 filters to have the closest RFs to the center of the image .
|
A central goal in systems neuroscience is to understand how large populations of neurons work together to enable us to sense , to reason , and to act . To go beyond single-neuron and pairwise analyses , recent studies have applied dimensionality reduction methods to neural population activity to reveal tantalizing evidence of neural mechanisms underlying a wide range of brain functions . To aid in interpreting the outputs of dimensionality reduction , it is important to vary the inputs to a brain area and ask whether the outputs of dimensionality reduction change in a sensible manner , which has not yet been shown . In this study , we recorded the activity of tens of neurons in the primary visual cortex ( V1 ) of macaque monkeys while presenting different visual stimuli . We found that the dimensionality of the population activity grows with stimulus complexity , and that the population responses to different stimuli occupy similar dimensions of the population firing rate space , in accordance with the visual stimuli themselves . For comparison , we applied the same analysis methods to the activity of a recently-proposed V1 receptive field model and a deep convolutional neural network . Overall , we found dimensionality reduction to yield interpretable results , providing encouragement for the use of dimensionality reduction in other brain areas .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neural",
"networks",
"engineering",
"and",
"technology",
"signal",
"processing",
"vertebrates",
"social",
"sciences",
"neuroscience",
"animals",
"mammals",
"primates",
"multivariate",
"analysis",
"mathematics",
"statistics",
"(mathematics)",
"white",
"noise",
"computational",
"neuroscience",
"vision",
"old",
"world",
"monkeys",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"monkeys",
"animal",
"cells",
"mathematical",
"and",
"statistical",
"techniques",
"principal",
"component",
"analysis",
"macaque",
"cellular",
"neuroscience",
"psychology",
"cell",
"biology",
"single",
"neuron",
"function",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"sensory",
"perception",
"computational",
"biology",
"physical",
"sciences",
"amniotes",
"statistical",
"methods",
"organisms"
] |
2016
|
Stimulus-Driven Population Activity Patterns in Macaque Primary Visual Cortex
|
In fungi , sexual identity is determined by specialized genomic regions called MAT loci which are the equivalent to sex chromosomes in some animals and plants . Usually , only two sexes or mating types exist , which are determined by two alternate sets of genes ( or alleles ) at the MAT locus ( bipolar system ) . However , in the phylum Basidiomycota , a unique tetrapolar system emerged in which four different mating types are generated per meiosis . This occurs because two functionally distinct molecular recognition systems , each encoded by one MAT region , constrain the selection of sexual partners . Heterozygosity at both MAT regions is a pre-requisite for mating in both bipolar and tetrapolar basidiomycetes . Tetrapolar mating behaviour results from the absence of genetic linkage between the two regions bringing forth up to thousands of mating types . The subphylum Pucciniomycotina , an early diverged lineage of basidiomycetes encompassing important plant pathogens such as the rusts and saprobes like Rhodosporidium and Sporidiobolus , has been so far poorly explored concerning the content and organization of MAT loci . Here we show that the red yeast Sporidiobolus salmonicolor has a mating system unlike any previously described because occasional disruptions of the genetic cohesion of the bipolar MAT locus originate new mating types . We confirmed that mating is normally bipolar and that heterozygosity at both MAT regions is required for mating . However , a laboratory cross showed that meiotic recombination may occur within the bipolar MAT locus , explaining tetrapolar features like increased allele number and evolution rates of some MAT genes . This pseudo-bipolar system deviates from the classical bipolar–tetrapolar paradigm and , to our knowledge , has never been observed before . We propose a model for MAT evolution in the Basidiomycota in which the pseudo-bipolar system may represent a hitherto unforeseen gradual form of transition from an ancestral tetrapolar system to bipolarity .
In fungi , sexual reproduction systems have been well characterized at the molecular level in numerous species belonging to the Ascomycota and the Basidiomycota , revealing striking similarities with sex determining regions in animals and plants [1] along with unique features such as the occurrence of thousands of sexual identity ( mating ) types in a single species [2] . Most fungi capable of sexual reproduction are heterothallic [3] which means that mating occurs only between distinct haploid individuals with compatible mating types . However , homothallic ( self-fertile ) mating is not unusual and several different genetic mechanisms have been shown to form the basis for this sexual behaviour [4]–[7] . Sexual reproduction systems in fungi have served as invaluable eukaryotic models of development [8] , transcriptional regulation [9] , [10] and signalling pathways [11] but research in the field has also been spurred by the fact that for many plant or animal pathogens , sexual reproduction is intimately linked with virulence ( e . g . the human pathogens Cryptococcus neoformans and Candida albicans [12] or the maize smut Ustilago maydis [13] ) . The chromosomal regions that determine the mating type in fungi are called the MAT loci and vary extensively in length and in genetic content in different fungal lineages . In most species within the Ascomycota , the two MAT loci in sexually compatible haploid individuals are short and contain completely unrelated sequences ( idiomorphs ) encoding transcription factors that regulate post-mating sexual development . These transcription factors belong to the alpha-domain , homeodomain or HMG families [14] . In the phylum Basidiomycota , mating of compatible haploid partners – homokaryotic hyphae or haploid yeast cells – originates a dikaryotic filamentous stage where subsequently basidia and basidiospores ( meiospores ) are formed [15] . Pioneering work on the mushroom Schizophyllum commune provided the first glimpse of a unique sexual compatibility system , the so-called tetrapolar or bifactorial system [16] . This system is only found in basidiomycetes because , unlike all other fungi studied so far , they possess two independent molecular determinants of mating type [17] . Sexual compatibility is usually determined at a first level by lipopetide pheromones and plasma membrane pheromone receptors that mediate cell-cell recognition , with the exception of some homobasidiomycetes ( the lineage that includes the mushrooms ) where fusion of homokaryotic hyphae is not restrained by pheromone-mediated interactions [2] , [18] . After cell fusion , progression through the sexual cycle requires overcoming a second compatibility hurdle that relies on a heterodimeric homeodomain transcription factor ( HD1/HD2 ) , encoded by a pair of divergently transcribed , closely linked genes [3] , [19]–[21] . The HD1/HD2 heterodimer is a transcriptional regulator of post-mating sexual development that is only active in the dikaryon because dimerization is restricted to subunits that originate from genetically different individuals . Dimerization of HD1 and HD2 proteins encoded by the same gene pair is prevented in haploid individuals by a self/non-self recognition mechanism that has been well characterized in mushrooms [22]–[24] and in smut fungi [25] , [26] . These studies showed that the self/non-self recognition domain resides on the highly variable N-terminal portion of the proteins . Accordingly , in HD1 proteins of the mushroom Coprinus cinereus , this domain was found to have exceptionally high evolution rates [27] . Almost without exception , the two classes of genes ( encoding pheromone/pheromone receptors and HD1/HD2 transcription factors ) are part of the MAT locus in basidiomycetes , which is the only genomic region for which the sequence varies according to the mating type . A notable aspect of the tetrapolar system is that it is multiallelic with respect to at least one of the two genetically unlinked classes of MAT genes , resulting in up to thousands of different mating types in some species [2] , [3] , [16] , [28] , [29] . Bipolar systems are , on the contrary , usually biallelic for both classes of genes [30] , [31] . Tetrapolar species can be found in all three major lineages of Basidiomycota interspersed with bipolar species [16] , [32] and , most significantly , do not seem to be uniformly distributed . In the subphylum Agaricomycotina , which encompasses the mushrooms , the majority of the species seems to be tetrapolar [16] . The subphylum Ustilaginomycotina harbours many plant pathogenic species like U . maydis , most of which exhibit tetrapolar mating systems very similar to those found in the Agaricomycotina . A substantial body of evidence [3] , [30] , [33] suggests that in these two subphyla , the few bipolar species were derived from tetrapolar ancestors as a result of coalescence between the regions encoding the two classes of MAT genes , like in the human pathogens Cryptococcus neoformans [30] and Malassezia globosa [34] and the barley smut Ustilago hordei [31] . In other cases ( the mushrooms Coprinellus disseminatus and Pholiota nameko ) the transition to bipolar behaviour was due to loss of the association of the pheromone receptor to the MAT locus [35] , [36] . The third subphylum , the Pucciniomycotina , contains about one-third of all species of Basidiomycota [37] . The majority of the species in the Pucciniomycotina are rust fungi , a large group of economically important obligate plant parasites . Classical mating studies indicate a predominance of bipolar systems in this subphylum both for rusts [38] , other plant parasites like Microbotryum violaceum [39] and for saprobic yeasts of the genera Rhodosporidium and Sporidiobolus [40] , [41] . However , detailed molecular analyses elucidating MAT gene content and function are very limited except for M . violaceum , in which size dimorphic sex chromosomes were identified and preliminarily characterized [42] and the pheromone receptor genes of both mating types were recently identified [43] . Since molecular phylogenies and ultrastructure suggest that Pucciniomycotina diverged first among subphyla of Basidiomycota [44] , the elucidation of the genetic structure of the mating systems in this lineage is of major importance for understanding the origin and evolution of basidiomycete MAT loci . The lack of knowledge on the sexual reproduction mechanisms of the Pucciniomycotina is mainly a consequence of the obligate parasitism of the rust fungi that seriously constrains experimental manipulation . This problem can be circumvented by studying related saprophytic organisms like those of the genera Rhodosporidium and Sporidiobolus since they are capable of completing their life cycle in culture . This led us to the recent characterization of one of the pheromone receptor regions in several bipolar red yeast species , including the heterothallic red yeast Sporidiobolus salmonicolor [45] . Here we report on the molecular characterization of the mating system in S . salmonicolor . We identified , for the first time in the Pucciniomycotina , the divergently transcribed genes encoding the HD1 and HD2 transcription factors and we investigated how HD1/HD2 transcription factors and the pheromone receptor system interact to produce the bipolar mating behaviour observed in S . salmonicolor . A new , non-bipolar non-tetrapolar , sexual mechanism was unveiled , in which occasional disruptions of the genetic cohesion of the bipolar MAT locus originate new mating types in a process that parallels that of tetrapolar systems .
We recently identified MAT A1 ( see the Methods section for changes in mating type designations in S . salmonicolor ) pheromone receptor and pheromone precursor genes in the heterothallic bipolar red yeasts Rhodosporidium toruloides and Sporidiobolus salmonicolor and in Sporobolomyces roseus [45] . These studies failed to uncover HD1/HD2 transcription factor genes in the vicinity of the pheromone receptor in any of the species studied . However , HD1 and HD2 homologs [32] , [45] were found in the completely sequenced genome of Sporobolomyces roseus , but at a large distance from the pheromone receptor gene . To get some insight in the genetic structure of MAT loci in red yeasts , we set out to characterize the MAT genotype of a set of 36 S . salmonicolor strains . This species was chosen because of previous studies concerning its mating behaviour [46] and of the availability of a 3× genome coverage in the form of Trace Archive sequences for one MAT A1 strain , considerably facilitating the identification of novel genes . We first identified the pheromone receptor gene in the complementary mating type of S . salmonicolor ( MAT A2 ) . Degenerate primers based on the sequences available for homologous MAT A2 receptors , including that of the closely related Rhodosporidium babjevae for which Trace Archive genomic sequences are available , failed to amplify the S . salmonicolor gene . An alternative approach was conceived , based on our previous observation of a high degree of synteny between the same mating type in different red yeast species [45] . For this , the region surrounding the pheromone receptor gene in R . babjevae was assembled stepwise using Trace Archive sequences and was subsequently scrutinized for the presence of genes exhibiting a higher degree of conservation across species than pheromone receptor genes . Two putative genes encoding an LSm-like protein ( LSm7 ) and a ribosomal protein L6 ( RibL6 ) were found respectively upstream and downstream of STE3 . A2 in R . babjevae . Sequences homologous to these genes were readily found by BLASTN search in the S . salmonicolor Trace Archives but they were differently organized , since the sequenced strain ( CBS 483 ) belongs to the opposite mating-type ( MAT A1; Figure S1 ) . PCR primers based on the S . salmonicolor LSm7 and RibL6 genes finally allowed the amplification of the intervening S . salmonicolor STE3 . A2 gene ( GenBank accession number GU474641 ) , attesting synteny conservation between S . salmonicolor and R . babjevae in this region within the same mating type . As expected , the predicted amino acid sequence of the S . salmonicolor receptor exhibited the highest similarity with the sequences of the R . babjevae ( 54% ) and M . violaceum ( 47% ) homologues . The divergently transcribed genes encoding the HD1 and HD2 transcription factors were identified in the Trace Archives sequences of S . salmonicolor using the homologous sequences of S . roseus . Primers based on the homeodomain conserved region of these genes amplified a fragment encompassing the highly variable 5′ regions of both genes as well as the intergenic region between the HD1 and HD2 genes ( Figure S1 ) . Homology to the HD1 and HD2 proteins from S . roseus was , as expected , limited to the homeodomain region since all functional HD proteins characterized so far were found to have highly variable N-terminal domains , even when different HD1/HD2 alleles from the same species are compared . Heterothallic red yeast species within the order Sporidiobolales were all described as having a bipolar behaviour in standard mating tests [40] , [41] . Therefore , it was striking to notice that in S . roseus , the distance between the pheromone receptor and HD1/HD2 regions although still not accurately determined , was in any case larger than 800 Kb [32] , [45] . Hence , if both classes of genes were part of a bipolar MAT locus with suppression of recombination over its entire length , it would be the largest such locus characterized so far in basidiomycetes . Such a region would nevertheless functionally resemble the MAT loci of C . neoformans [30] , [47] and U . hordei [31] , [48] . Alternatively , one of the compatibility check points might no longer be required to determine sexual identity , as previously observed for two mushroom species [35] , [36] . In light of this , it was important to investigate how HD1/HD2 transcription factors and the pheromone receptor system interacted to produce the bipolar mating behavior . Isolation of both classes of MAT genes in S . salmonicolor enabled us to examine this question for the first time . We first used specific PCR primers to assess the correlation between the presence of the alternate pheromone receptor genes ( STE3 . A1 and STE . A2 ) and mating behavior , in 36 natural isolates of S . salmonicolor ( Table S1 ) . Without exception , STE3 . A1 was present in MAT A1 strains , whereas the STE3 . A2 receptor gene was found in MAT A2 strains . Hence , the pheromone receptor was clearly associated to mating behaviour . Next , PCR fragments encoding the 5′ portions of the HD1and HD2 genes were obtained from the 36 S . salmonicolor strains of both mating types and sequenced ( GenBank accession numbers GU474649–GU474693 ) . Surprisingly , instead of two HD1/HD2 alleles , each strictly linked to one of the mating types , as might be expected for a bipolar species , 13 HD1/HD2 alleles exhibiting substantial sequence divergence were uncovered ( Figure 1A ) . Such numbers of HD1/HD2 alleles were previously observed only for tetrapolar species [16] . In U . maydis , for example , 33 HD1/HD2 alleles were identified [49] which do not have a particular association to either one of the two pheromone receptors because the two MAT regions are located on different chromosomes [3] and thus segregate independently at meiosis . On the contrary , in the bipolar species characterized so far , only two HD1/HD2 alleles are present , each of which is linked to one of the two pheromone receptors forming two large bipolar loci [30] , [31] . Recombination between the two genetically linked MAT regions is suppressed in these species as a result of extensive sequence divergence , gene inversions and the accumulation of repetitive elements reminiscent of sexual chromosomes in animals and some plants [1] , [30] , [31] , [50] . In S . salmonicolor , each of the 13 alleles always appears associated with the same receptor , suggesting some form of genetic linkage between the two regions ( Figure 1 ) . However , each receptor is associated with seven ( Ste3 . A2 ) or six ( Ste3 . A1 ) different HD1/HD2 alleles which is in sharp contrast with other bipolar species ( Figure 1 ) . There is apparently no bias regarding the geographic distribution of HD1/HD2 alleles , since strains carrying the same HD1/HD2 allele were isolated from very diverse locations worldwide ( Table S1 ) . To shed some light on the phylogenetic relationship between the HD1/HD2 alleles , we also examined the alleles present in a very closely related species , S . johnsonii . Several lines of evidence suggest that S . salmonicolor and S . johnsonii are undergoing a speciation process in which pre-zygotic barriers are absent , since crosses between sexually compatible strains of the two species normally yield dikaryotic mycelium and teliospores [46] . Intriguingly , extant allele diversity seems to have been generated after the onset of this incipient speciation event , as HD1/HD2 alleles from the two species are , in general , phylogenetically distinct ( Figure S2 ) . However , some incongruities between rDNA and MAT gene phylogenies can be observed ( Figure S2 ) , indicative of a certain degree of gene flow between the two species . For example , S . salmonicolor strain CBS 2634 has a S . johnsonii rather than a S . salmonicolor receptor allele and , in addition , a group of five S . salmonicolor strains carry HD1/HD2 alleles ( A1–5 , A2–16 , A1–6; Figure S2 ) which seem to be phylogenetically more related to the S . johnsonii clade . The intraspecific common ancestry of HD1/HD2 alleles contrasts with the polymorphism observed for the pheromone receptor genes . The latter genes were found to exhibit a clear and ancient trans-specific polymorphism ( Figure S3 ) , as previously reported for M . violaceum [43] . Hence , the relationship between mating type and HD1/HD2 allele number and distribution is unexpectedly complex in S . salmonicolor , and the possibility that the HD1/HD2 transcription factors might have recently lost their linkage to MAT and are no longer involved in determining sexual compatibility in this species could not be readily discarded . To examine this , we determined first the evolution rates ( dN/dS ) of nine HD1 alleles of S . salmonicolor and S . johnsonii ( GenBank accession numbers GU474694–GU474702 ) . Prior studies in the mushroom Coprinus cinereus brought to light exceptionally high evolution rates for the domains in HD1 proteins involved in self/non-self recognition [27] . We observed even higher evolution rates for the homologous domains in the HD1 proteins of S . salmonicolor ( >1 in some segments , indicative of adaptive selection; Figure 2 and Figure S4 ) , which constituted the first evidence that these transcription factors contribute to determine sexual compatibility in S . salmonicolor , together with the pheromone receptor system . We sought further for an explanation of how the “one-to-many” ratio between pheromone receptor and HD1/HD2 genes , characteristic of tetrapolar systems , results in bipolar behaviour with only two mating types in S . salmonicolor . To this end , we investigated the possibility that the cohesion of the bipolar MAT locus in S . salmonicolor might be occasionally disrupted , giving rise to new receptor/HD allele combinations . Such recombination events would explain the common ancestry of extant A1- and A2-linked HD1/HD2 alleles ( Figure 1A ) and indeed the fact that all HD1/HD2 alleles in the species seem to share a common ancestor more recent than the formation of the species itself ( Figure S2 ) . Therefore , bipolarity in S . salmonicolor could be a consequence of the scarcity of recombination events on the one hand and of the exceptionally high evolution rates of HD1/HD2 proteins on the other hand . The combination of these two factors would lead to the association of each extant HD1/HD2 allele with only one of the alternate receptors , the hallmark of bipolarity . Alternatively , it could not be readily excluded that the observed apparent linkage could be due to population genetic phenomena related to the frequency of sexual reproduction and population size . To address the possibility of partial genetic linkage between the two MAT regions , we studied in detail the latter stages of the sexual cycle ( germination of teliospores with the formation of basidia and basidiospores; Figure S5 ) in one cross between strains ML 2241 ( MAT A1 ) and CBS 6832 ( MAT A2 ) . The progression through meiosis was microscopically monitored showing that although four nuclei could be observed after meiosis , only two basidiospores were formed , each arising from one of the two basidial compartments ( Figure S5B ) . The two basidiospores were always found to be binucleate and they typically exhibited asynchronous germination ( Figure S5 ) . The molecular mating types of the progeny of eight meioses from the same cross were examined by micromanipulation of teliospores in the initial stages of germination . Several colonies isolated from each germinating teliospore were analysed and , in only one of the eight events examined ( T1 ) , were the two versions of the parental mating type genes recovered ( Figure 3 ) . Our interpretation is that the asynchronous germination of the two basidiospores results in a very strongly biased composition of the colony towards descendants of the basidiospore that germinates first . Micromanipulation of the basidiospores instead of the germinating teliospore would probably be required to recover the germination products of both basidiospores with higher frequency , but this proved to be technically very difficult due to the fact that germination takes place inside the agar and the structures to be manipulated are friable . Furthermore , the various colonies examined for each teliospore are most likely mitotic clones , as they invariably shared the same parental allele of the DMC1 gene , ( Figure 3 , GenBank accession numbers HM133872–HM133874 ) . In six of the meiotic events examined , linkage between the pheromone receptor and the HD1/HD2 regions was maintained , as would be expected in a bipolar system ( Figure 3 ) . Germination of one additional teliospore yielded a strain which was apparently diploid since it carries pheromone receptor and HD1/HD2 alleles from both parental strains ( Figure 3 ) and has a self-fertile phenotype with the formation of mycelium and teliospores . Self-fertile and asexual strains can be found among natural isolates of red yeasts of the Sporidiobolales [40] , [41] and the origin of the first may be related to the occasional formation of diploid yeast strains . In addition to our present observations , the formation of diploid strains was also previously reported in R . toruloides [51] . However in none of the species studied so far are these diploid states prevailing , leading to the presumption that although these strains do not exhibit obvious growth defects , their genetic makeup may carry some selective disadvantages . Moreover , our preliminary observations suggest that teliospores formed by the S . salmonicolor diploid strain do not germinate , which argues against the co-existence of homothallic and heterothallic life styles in this species . Interestingly , the type strain of S . johnsonii ( CBS 5470T ) may have originated from a diploid strain because it carries both pheromone receptor alleles . However , only one HD1/HD2 gene pair is present which may indicate that a self-compatible HD1/HD2 gene pair was generated by recombination between the two parental alleles ( Figure S2 ) . The type strain of S . johnsonii is homothallic [41] , [46] but does not form basidia , differing in that respect from all the other strains of this species , which carry only the MAT A1 receptor gene and are in general capable of mating with S . salmonicolor MAT A2 strains . One germinated teliospore yielded the most striking result , since it produced an haploid strain in which recombination occurred , causing the HD1/HD2 allele of parental strain CBS 6832 ( MAT A2 ) to become associated with the Ste3 . A1 receptor ( Figure 3 ) , a combination that is not found among natural isolates ( Figure 1A ) . The new mating type ( strain T7 ) was unable to complete the sexual cycle when crossed with either parental strain ( Figure 4 ) . A cross with the parent sharing the same pheromone receptor gene resulted in a complete failure to switch to filamentous growth , whereas homozygosity at the HD1/HD2 region precluded progression through the sexual cycle , but allowed some pseudohyphal growth , probably as a result of cell-cell pheromone signalling ( Figure 4 ) . This demonstrates unequivocally that both compatibility regions are required for sexual reproduction in S . salmonicolor . The new mating type did not show obvious defects in vegetative growth , exhibiting growth rates in synthetic and complete media similar to those of the parental strains ( results not shown ) . It also had normal sexual proficiency when crossed with strains carrying different alleles at both MAT regions ( Figure 4 ) , forming dikaryotic mycelium with teliospores that germinate normally ( result not shown ) . Recombination between the two classes of MAT genes seems to be possible but infrequent , a situation that , to our knowledge , has never been described before in basidiomycetes and can be regarded as intermediate between bipolar and tetrapolar systems . Therefore , this observation called for a closer examination of the meiotic events described above . In particular , it was important to establish how genetic markers unrelated to MAT segregated in these meiotic events , thereby getting some insight in the recombination frequency characteristic of this species in autosomal regions while simultaneously confirming that a normal meiosis had taken place in all the cases examined . It should be noted that a more precise determination of recombination frequencies and a statistically significant demonstration of linkage between genomic regions would require the examination of a larger number of meiotic products , but nevertheless examination of the eight available strains yields significant information concerning recombination inside and outside the MAT region . To characterize the haploid progeny strains , we chose nine genes located in four different S . roseus scaffolds ( Figure 5 ) . Pairs of genes located on the same scaffold were at distances ranging between 0 . 525 and 1 . 87 Mb from each other ( Figure 5B ) . The selected genes were partially amplified in the two parental strains used in this cross ( ML 2241 and CBS 6832 ) and sequence polymorphisms were scored which allowed the two parental alleles to be distinguished and their fate after meiosis to be tracked ( GenBank accession numbers HM133857–HM133883 ) . Sequencing of this set of genes in the seven haploid progeny strains confirmed that these markers segregate independently of MAT-specific genes ( Figure 5 ) . Markers located more than 1 . 2 Mb apart seem to be genetically unlinked , while the results suggested partial linkage for genes that are 525 to 920 kb apart , with an average of 32 kb/cM for these autosomal regions . The scaffolds containing the two MAT regions were subsequently subjected to a similar analysis . In scaffold 7 , five genes , including the HD1/HD2 pair were found to contain polymorphisms and were studied with respect to their parental origin in the seven haploid progeny strains ( GenBank accession numbers GU474757–GU474780 , HM133833–HM133856 ) . In the scaffold harbouring the STE3 gene , the parental origin of a total of eight additional genes exhibiting polymorphisms was determined ( GenBank accession numbers GU474709–GU474756 , HM133785–HM133832 ) . The relative orientation of the two scaffolds could not be previously established [45] , but the results of our scrutiny of meiotic recombination in these regions suggest linkage between the IsocL gene in scaffold 7 and the PAN6 gene in scaffold 9 . This implies that the relative orientation of the two scaffolds is most likely as depicted in Figure 5A . The analysis of the markers in these regions brought to light , in addition to sites where probably crossovers occurred , several instances where very likely gene conversion took place instead , one of which curiously involves the HD1/HD2 gene pair ( Figure 5A ) . In line with this , we found that the phylogeny of several genes located in the two scaffolds harbouring MAT genes also occasionally denotes signs of past gene conversion events when multiple strains of S . salmonicolor and S . johnsonii are examined ( Figure S6 ) . For example , the LSm7 gene located close to STE3 , exhibits an unusual mosaic phylogeny , which is species-specific in most of the gene but is MAT specific at the 3′-end , proximal to STE3 ( Figure S6 ) . This suggests that this gene may have integrated the MAT locus long ago as might be expected from its close proximity to a key MAT-specific gene , but that the 5′end underwent conversion more recently , after incipient separation of the two species . Phylogenetic analysis of the RibL18ae , RNAPOL , NGP1 and AKOR2 genes also suggests gene conversion events in S . salmonicolor strain CBS 1012 ( Figure S6 ) . These observations support the idea that gene conversion is an important mechanism to maintain species-specific sequences at the S . salmonicolor MAT locus . This may , in turn , reflect the need to counteract the effect of recombination suppression which could lead to mating type-linked deleterious mutations . It is also important to note that sequence divergence between mating types of genes located in the vicinity of the receptor/pheromone MAT region is much lower in the red yeasts examined than , for example in Cryptococcus neoformans ( Figure S6 , [47] ) , although in the later species some events of gene conversion within the MAT loci have also been noticed [47] . Three crossovers were mapped to a region of 27 kb adjacent to the HD1/HD2 gene pair ( T5 , T6 and T7; Figure 5 ) , which seems to configure a hotspot for recombination ( 0 . 63 kb/cM ) when compared with the average frequency of recombination observed for the autosomal regions examined ( ∼32 kb/cM ) . Interestingly , hotspots have been mapped close to the borders of the MAT locus in Cryptococcus neoformans [52] and were proposed to play an important role in the evolution of these specialized genomic regions . Three recombination events were also detected on the opposite side of the putative MAT locus in a region of ∼220 kb ( T1 , T2 and T3; Figure 5 ) but in this case only one genetic marker ( MIP ) was studied on one side of this position , which makes it difficult to distinguish between gene conversion and crossover events . Finally , two crossovers were detected in the proximity of the PAN6 gene , one of which was the cause of the only instance of recombination detected between the STE3 and the HD1/HD2 genes ( T7 , Figure 5 ) . In the other case ( T4 , Figure 5 ) , the MAT A1 parental mating type prevailed , due to apparent gene conversion involving the HD1/HD2 gene pair . In the 1 . 2 Mb region between the HD1/HD2 and PAN6 genes no crossovers seem to have occurred , suggesting that recombination frequency is lower than average or possibly suppressed in this region . Hence , we propose that in S . salmonicolor a new type of fungal sex determining region operates , which is neither tetrapolar nor strictly bipolar . S . salmonicolor seems rather to have a large pseudo-bipolar MAT locus in which recombination in the region between the two MAT regions is infrequent but not suppressed , allowing for occasional disruptions of its genetic cohesion . The proposal of an intermediate system between bipolar and tetrapolar is based mainly on two kinds of observations , which are in line with each other . Firstly , the allelic distribution and number of MAT genes depicted in Figure 1 outlines a situation that is itself intermediate between bipolar and tetrapolar ( multiple HD1/HD2 alleles but only two mating types ) . Secondly , when trying to clarify this new finding , we found that recombination may occur in the intervening region between the two MAT regions , which provides a likely and plausible explanation for the observed resemblance with tetrapolar systems in what concerns evolution and number of HD1/HD2 alleles . On the other hand , the genomic region around the receptor/pheromone locus exhibiting synteny breaks , gene inversions and sequence divergence between the two mating types is significantly larger in S . salmonicolor than in tetrapolar species with similar numbers of MAT gene alleles , like U . maydis ( Figure S1 , [47] ) . We envisage that the offspring of the relatively rare events of meiotic recombination in the MAT locus of S . salmonicolor can be rescued from reproductive isolation by the existence of multiple HD1/HD2 alleles associated to each of the two alternate receptors in natural populations . In such a pseudo-bipolar system , negative frequency-dependent selection could operate to preserve HD1/HD2 allele diversity and explain the exceptionally high evolution rates of HD1 proteins . Both the content and the organization of the basidiomycete MAT loci are unlike all other fungi , because compatibility is encoded in the MAT locus itself , while in other fungi the genes present in the MAT locus regulate expression of pheromones and pheromone receptors that are encoded elsewhere in the genome . The available data does not suffice to address definitively the question of how the basidiomycete MAT locus arose , but one hypothesis has been recently put forward that postulates that the pheromone receptor and/or the HD systems may have evolved in a self compatible manner [49] . The emergence of self-incompatible alleles by mutation or recombination could have subsequently laid the basis for the emergence of the tetrapolar system . There is currently insufficient evidence , also including our present observations , to state decisively whether the ancestral basidiomycete mating system was initially bipolar or tetrapolar , at the onset of the involvement of both the HD and pheromone receptor systems in determining mating specificity . However , our observations concerning the mating system in S . salmonicolor , suggest that novel and useful insights may be gained in this respect by a wider phylogenetic sampling of mating systems in the Pucciniomycotina . In line with this , we subsequently looked into the HD1/HD2 allelic distribution of Rhodosporidium babjevae , another bipolar red yeast species phylogenetically related to S . salmonicolor . Also in this case , multiple HD1/HD2 alleles were associated with each receptor ( Figure S7 ) . Since this allelic distribution is the hallmark of the pseudo-bipolar system , we conclude that rather than being an oddity restricted to S . salmonicolor , this system is likely to operate in many of the yeast species of the order Sporidibolales [53] and probably also in other members of the subphylum Pucciniomycotina that were thus far thought to be bipolar . The most parsimonious explanation for this is , in our view , that a very similar system was present in the common ancestor of Rhodosporidium and Sporidiobolus . This implies that pseudo-bipolarity it is not a short-lived transitional mating system , like those presumed to have marked the transition to the bipolar state in C . neoformans [30] , [31] and U . hordei [3] . Rather , it suggests that it may be evolving from an ancestral system in which the two classes of MAT genes were located on the same chromosome , but of which it is not possible to infer from currently available data whether the two MAT regions were genetically linked . The model shown in Figure 6 depicts the possibility that the ancestral basidiomycete mating system may have been similar to the pseudo-bipolar system found in S . salmonicolor . We highlighted this possibility among others , because the Pucciniomycotina is the earliest derived basidiomycete lineage and seems therefore more likely that members of this group have retained mating systems akin to the common ancestral system . In this scenario , the Pucciniomycotina would have diverged before the stabilization of the tetrapolar system ( with the two MAT regions on different chromosomes ) , which is presumed to be the ancestor of the mating systems in the other two lineages ( Ustilaginomycotina and Agaricomycotina ) [49] . The model proposed in Figure 6 depicts this possibility and accommodates the possible occurrence of abrupt as well as gradual transitions to bipolar mating behaviour , the first but not the second being triggered by gross genomic rearrangements . An intermediate , pseudo-bipolar mating system entails odds of inbreeding which are in between those observed for the bipolar and tetrapolar systems ( 50% and 25% , respectively; [33] ) . This may provide for some species the right balance between genetic stability ( imparted by inbreeding ) and variation ( favoured by outcrossing ) and may , in turn , delay the attainment of a strictly bipolar state if a system is drifting away from a tetrapolar configuration without abrupt occurrences , such as large chromosomal rearrangements . Our time course microscopic observation of meiotic events in S . salmonicolor indicates that basidiospores exhibit asynchronous germination ( Figure S5C ) , decreasing the probability of a cross between siblings immediately after meiosis . This , together with the occasional meiotic recombination events , would generally decrease the odds of selfing when compared with a typical bipolar system , like that of the phylogenetically related plant pathogen Microbotryum violaceum [39] . Hence , the pseudo-bipolar mating system emerges as a remarkable novel context in which to explore how life-style , ecology and modes of reproduction interplay in the evolutionary history of eukaryotes .
The list of strains studied and relevant information associated to them is given in Table S1 . We re-assigned mating type designations of S . salmonicolor in order to match the “molecular mating type” identified by PCR detection of the pheromone receptor alleles ( STE3 . A1 or STE3 . A2 ) . Preparative PCR reactions were performed in a final volume of 50 µl with the following components ( unless stated otherwise ) : 2 mM of MgCl2 , 0 . 20 mM of each of the four deoxynucleoside triphosphates ( GE Healthcare ) , 1% DMSO , 0 . 8 µM of each primer , 100 ng of genomic DNA , and 1 U Taq DNA polymerase ( Fermentas , Canada ) . Thermal cycling consisted of a 5-minute denaturation step at 95°C , followed by 35 cycles of denaturation at 94°C for 30 s , 30 s at the annealing temperature ( variable ) , and extension at 72°C ( variable time ) . For annealing temperatures , extension times and primer sequences see Table S2 . A final extension of 7 min at 72°C was performed at the end of each reaction . The pheromone receptor gene ( pr-MatA2 ) from Microbotryum violaceum ( GenBank accession number EF584741 ) was used to perform a BLASTN search in the NCBI Trace Archive sequences of the genome project of Rhodosporidium babjevae strain WP1 ( MAT A2; www . jgi . doe . gov/sequencing/statusreporter/psr . php ? projectid=52130 ) . The complete STE3 . A2 sequence of R . babjevae , in addition to the flanking regions , were assembled ( Table S3 ) and subsequently scrutinized for the presence of genes exhibiting a higher degree of conservation across species than pheromone receptor genes . Sequences of two putative genes encoding an LSm-like protein ( LSm7 ) and a ribosomal protein L6 ( RibL6 ) were used to design a pair of degenerate primers ( MC122 and MC123 ) to amplify and sequence the intervening region in S . salmonicolor MAT A2 strain CBS 490T . This region contained the STE3 . A2 S . salmonicolor gene , which was sequenced by primer walking using primers MC126 and MC127 ( Figure S1 ) . Using the sequence of the HD1 gene of Sporobolomyces roseus [45] , a BLASTN search was performed in the NCBI Trace Archive database of S . salmonicolor and positive hits ( Table S3 ) were assembled into a complete HD1 gene . The upstream flanking region of the HD1 was assembled stepwise and inspected using AUGUSTUS software [54] , revealing the presence of a divergently transcribed HD2 homologue . The deduced protein sequences of HD1 and HD2 of S . roseus and S . salmonicolor were aligned and the conserved regions were used to design specific primers ( MC103 and MC104 ) to amplify and sequence the corresponding N-terminal and intergenic regions of the HD1/HD2 genes in all S . salmonicolor and S . johnsonii strains ( Figure S1 , Table S3 ) . To confirm mating behavior and sexual compatibility , 2–4 day-old cultures were crossed on corn meal agar ( Difco ) , incubated at room temperature for 1 week , and examined microscopically using phase-contrast optics for production of mycelium with clamp connections and teliospores . Diagnostic PCR with primers for STE3 . A1 ( MC053 and MC054 ) and STE3 . A2 ( MC126 and MC127 ) were carried out to identify the pheromone receptor genes . Strains CBS 6832 ( A2–6 ) and ML 2241 ( A1–3 ) were mixed on corn meal agar ( Difco ) and incubated at room temperature ( ∼22°C ) for 2 weeks , to allow for abundant production of teliospores . Small ( ∼0 . 5 cm ) agar blocks containing teliospores were soaked in sterile distilled water for 8–10 weeks at 4°C . After this resting period , a suspension of teliospores was obtained by gently mashing the agar blocks with a small pestle . Teliospore germination was induced by transferring this suspension to 2% low melting point ( LMP ) -agarose plates incubated at room temperature for 2 weeks . To follow the germination of selected teliospores , small drops of the same suspension were transferred to 2% LMP-agarose-coated-slides . The different stages of meiosis were observed microscopically by staining several germinating teliospores with Safranin O [55] . Observations were made daily with 100× magnification using a Leica DMR microscope equipped with brightfield and differential interference contrast optics and microphotographs were recorded using a Leica DFC320 digital camera ( Leica Microsystems GmbH , Wetzlar ) . Teliospores resulting from the cross of strains CBS 6832 ( A2–6 ) and ML 2241 ( A1–3 ) were germinated as described above . Using a micromanipulator , teliospores in the initial stages of germination , i . e . with a non-septate basidium initial , were individually separated , transferred to MYP agar [46] and incubated at room temperature . The colony that formed after 1–2 days was re-streaked to obtain colonies derived from single cells . In three to six of these colonies , segregation of the two MAT regions was assessed by diagnostic PCR with specific primers for STE3 . A1 ( MC053 and MC054 ) , STE3 . A2 ( MC126 and MC127 ) and HD1/HD2 ( MC103 and MC104 ) . The HD1/HD2 alleles were discriminated after amplification by digestion with restriction enzyme Rsa I . DNA and protein sequences were aligned with ClustalW 1 . 81 [56] with minor manual corrections . For the phylogeny of HD1/HD2 alleles , FindModel ( http://www . hiv . lanl . gov/content/sequence/findmodel/findmodel . html ) , which is an implementation of ModelTest [57] , with the Akaike information criterion ( AIC ) was used . The Tamura-Nei ( TrN ) +G model [58] , [59] ( shape parameters 0 . 98049 and 1 . 08652 in Figure 1 and Figure S2 datasets , respectively ) was employed in MEGA4 [60] using the Neighbor-Joining algorithm [61] and bootstrap values from 1000 replicates . For the phylogeny of the pheromone receptors ( Figure S3 ) the substitution model of protein evolution was selected using ProtTest [62] with AIC . The WAG+I+G [59] , [63] , [64] model ( shape parameter = 2 . 99; proportion of invariable sites = 0 . 035 ) was employed in PHYML [65] using maximum likelihood and bootstrap values from 100 replicates . To obtain the sequence of the 3′end of HD1 gene the region encompassing the homeodomain and a conserved motif located 10 bp upstream of the STOP codon was amplified using primers MC111 and MC112 . The 5′end of the HD1 gene was amplified using primers MC103 and MC104 ( Figure S1 ) . Evolution rates were estimated by a Window Analysis of dN and dS , using the online interface of WINA 0 . 34 [66] , in a sliding window ( size = 20 ) along the alignment of nine HD1 alleles . Based on the available genomic information of the closely related species Sporobolomyces roseus , and assuming that synteny is maintained in S . salmonicolor [41] , eight genes from the pheromone receptor region ( scaffold 9 ) and four genes from the HD1/HD2 region ( scaffold 7 ) were selected . NCBI S . salmonicolor Trace Archive sequences of these genes were obtained ( Table S3 ) and assembled . Primers were designed for partial sequencing of these genes ( Table S2 ) . Protein-coding DNA sequences were deduced after removal of putative introns , either manually or using AUGUSTUS software [54] , automatically aligned using ClustalW and manually edited according to the superimposed amino acid sequences . Synonymous substitutions ( dS ) values and the divergence percentage between mating type-specific alleles were calculated using DnaSP 5 . 0 [67] for a set of 10 S . salmonicolor strains ( Figure S6 ) . Phylogenies were obtained in MEGA4 [60] using Maximum Parsimony ( MP ) , with the close-neighbor interchange algorithm for heuristic searches and bootstrap values from 1000 replicates . To detect differences among the meiotic progeny , we selected nine genes located on four different scaffolds ( 1 , 2 , 3 and 10; Figure 5 ) in the S . roseus genome . Sequences of these genes in S . salmonicolor were assembled from the NCBI Trace Archive ( Table S3 ) and primers were designed to amplify partial sequences in parental strains CBS 6832 and ML 2241 ( Table S2 ) . Sequence polymorphisms were scored and used to track the parental origin of the alleles in all the meiotic progeny . This analysis was performed either by PCR and sequencing or by RFLP analysis ( Table S2 ) . The genes previously used to obtain mating type-specific phylogenies were similarly assessed in this genotyping analysis .
|
Sexual reproduction in fungi is regulated by genomic regions called MAT loci that determine sexual identity in a manner comparable to that of sex chromosomes in animal and plants . In most fungi , sexual reproduction is bipolar , i . e . , two alternate and distinct sets of genes at the MAT locus determine two mating types ( the equivalent to sexes ) . In the Basidiomycota , which is the fungal lineage that includes the mushrooms , a unique ( tetrapolar ) sexual reproduction system evolved in which the mating type is determined by two functionally independent and genetically unlinked regions . In addition , the tetrapolar system functions with multiple alleles for at least one the two classes of MAT genes . This potentially generates thousands of mating types , as observed for many mushroom species . Here we report on the molecular characterization of the mating system in the basidiomycetous red yeast Sporidiobolus salmonicolor , which belongs to the Pucciniomycotina , the earliest derived lineage of the Basidiomycota that has remained virtually unexplored with respect to gene content and structure of MAT loci . Our results revealed for the first time a mating system that is neither tetrapolar nor strictly bipolar , and we propose a model for the evolution of basidiomycete MAT loci that accommodates this novel finding .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"molecular",
"biology/molecular",
"evolution",
"microbiology/microbial",
"evolution",
"and",
"genomics"
] |
2010
|
A Deviation from the Bipolar-Tetrapolar Mating Paradigm in an Early Diverged Basidiomycete
|
Retroviral insertional mutagenesis ( RIM ) is a powerful tool for cancer genomics that was combined in this study with deep sequencing ( RIM/DS ) to facilitate a comprehensive analysis of lymphoma progression . Transgenic mice expressing two potent collaborating oncogenes in the germ line ( CD2-MYC , -Runx2 ) develop rapid onset tumours that can be accelerated and rendered polyclonal by neonatal Moloney murine leukaemia virus ( MoMLV ) infection . RIM/DS analysis of 28 polyclonal lymphomas identified 771 common insertion sites ( CISs ) defining a ‘progression network’ that encompassed a remarkably large fraction of known MoMLV target genes , with further strong indications of oncogenic selection above the background of MoMLV integration preference . Progression driven by RIM was characterised as a Darwinian process of clonal competition engaging proliferation control networks downstream of cytokine and T-cell receptor signalling . Enhancer mode activation accounted for the most efficiently selected CIS target genes , including Ccr7 as the most prominent of a set of chemokine receptors driving paracrine growth stimulation and lymphoma dissemination . Another large target gene subset including candidate tumour suppressors was disrupted by intragenic insertions . A second RIM/DS screen comparing lymphomas of wild-type and parental transgenics showed that CD2-MYC tumours are virtually dependent on activation of Runx family genes in strong preference to other potent Myc collaborating genes ( Gfi1 , Notch1 ) . Ikzf1 was identified as a novel collaborating gene for Runx2 and illustrated the interface between integration preference and oncogenic selection . Lymphoma target genes for MoMLV can be classified into ( a ) a small set of master regulators that confer self-renewal; overcoming p53 and other failsafe pathways and ( b ) a large group of progression genes that control autonomous proliferation in transformed cells . These findings provide insights into retroviral biology , human cancer genetics and the safety of vector-mediated gene therapy .
The oncogenic potential of murine γ-retroviruses ( MLVs ) stems from proviral integration into host DNA , a mutagenic process which can result in activation or disruption of critical host cell genes [1] . Moreover , by sequential integrations in the nascent tumour cell , MLVs can drive multiple steps in the oncogenic process . These features have led to the use of MLVs as screening tools for genes relevant to cancer , particularly haematopoietic malignancies . The reach of this approach has grown considerably with the development of high throughput methods for cloning and sequencing analysis of host-virus junctions at insertion sites , facilitating screens of large tumour panels and identifying hundreds of genes of potential relevance to cancer . Importantly , genes identified by this method frequently map to orthologous sites of mutation in human cancer [2] , [3] . Moreover , retroviral insertional mutagenesis ( RIM ) provides a complementary approach to whole genome sequencing and copy number analysis in cancer , as RIM has the potential to uncover genes that are rarely mutated but more commonly subject to indirect processes including epigenetic modification [4] . Furthermore , large scale analyses of co-occurrence of target genes can identify patterns indicating collaborative or redundant relationships between cancer genes [5] , [6] . Despite the wealth of information provided by these studies , it is not yet known whether two events are sufficient for lymphoid transformation or whether higher order collaborations between more than two target genes are required . Target gene interactions can be explored functionally when combined with manipulation of the mouse genome and mice with an activated oncogene or mutant tumour suppressor gene in the germ-line often show accelerated tumour onset [7] , [8] . RIM tagging in this context reveals preferential targeting of specific collaborating genes , which can be confirmed by analysis of compound transgenic mice [1] . Moloney murine leukaemia virus ( MoMLV ) is an oncogenic γ-retrovirus that has been widely used in RIM studies [3] , [9] , [10] and owes its potency to a duplicated enhancer element in the proviral long terminal repeats ( LTRs ) [11] . Notably , the LTRs and backbone of this virus formed the basis of retroviral vector systems used in early trials of human gene therapy , where leukaemia resulting from insertional activation of host genes has been a significant adverse outcome [12] . In mice , the target genes for MoMLV that have been identified to date show a predominance of oncogene activation events over tumour suppressor disruption , consistent with the observed low rate of loss of heterozygosity in MoMLV lymphomas [13] . However , these findings presented a long-standing puzzle in light of the effect of germ-line inactivation of the major tumour suppressor p53 , which confers rapid onset T-cell lymphomas with a similar broad phenotypic spectrum to MoMLV but shows relatively weak cooperation with MoMLV [14] . We hypothesised previously that the MoMLV oncogenic programme must neutralise the tumour suppressor activity of p53 , circumventing the need for direct mutations in the pathway [14] , [15] . In support of this proposal we showed that the potent combination of two MoMLV target genes , Myc and Runx2 , could overcome the need for genetic inactivation of the p53 pathway , despite the fact that both oncogenes evoke p53 growth suppression and collaborate strongly with p53 deficiency [16] . Nevertheless , this combination still appears to be insufficient for full transformation , as double transgenic tumours emerge as clonal outgrowths from a polyclonal premalignant phase [17] . We showed previously that tumour onset could be accelerated by retroviral infection and a RIM screen identified a number of candidate third hit genes , including Pim1 , a gene that accelerates tumour onset when combined with MYC/Runx2 in the germ-line [9] , [18] . In this study we have conducted a further screen on the same progressing lymphomas , using a deep sequencing method ( splinkerette/454 ) which is orders of magnitude more sensitive than previous shotgun cloning methods . Sequencing at this depth raises another potential concern , as γ-retroviruses including MoMLV display preferential integration at transcriptional start sites and other chromatin feature that may also entail a bias towards proto-oncogenes [19]–[21] . However , we present multiple lines of evidence for post-integration selection as the dominant force shaping the progression ‘integrome’ . Moreover , we find that a surprisingly large fraction of the known MoMLV target gene spectrum is detectable in the integrome , indicating that any one among hundreds of genes can contribute to driving clonal outgrowth . However , there is a clear hierarchy of target genes that are selected from a large gene pool generated by the intrinsic preferences of γ-retrovirus integration . Another striking finding is the genetic bottleneck to transformation imposed by transgenic CD2-MYC , which is highly dependent on Runx gene activation . Comparison with other transgenic models of Myc over-expression shows that these each display potent selection from a small pool of master collaborating genes . These genes share the capacity to suppress the p53 pathway but are differentially recruited according to lymphoid lineage and developmental stage . The identification of a small gene set that confers the lymphoma initiating cell phenotype and is conserved in human disease has significant implications for targeted interventions .
Relevant features of the CD2-MYC and CD2-Runx2 transgenic mice are displayed in Figures 1 and S1 . The disease-free survival of most parental transgenic mice has been attributed to variegated expression under CD2 locus control region ( LCR ) control [22] along with counter-selection by failsafe processes [22] , [23] . As previously described [9] , [17] , [23] , co-expression of both transgenes results in rapid onset lymphomas in 100% of mice , but the tumours typically display a single predominant clone as illustrated by T-cell receptor gene rearrangement ( Figure S1 ) . Neonatal infection with MoMLV leads to accelerated lymphoma onset , increased clonal complexity and lymphoid dissemination , although the tumours retain the characteristic bimodal phenotype seen in the absence of infection ( CD8+ , CD4+/− , TCRhi ) [16] . Here , a panel of 28 lymphomas was analysed by RIM/DS ( splinkerette/454 ) . Processing of reads as described in Methods yielded 12 , 485 unique retroviral insertion sites ( RISs ) , compared to 272 by previous manual cloning and sequencing methods [9] . Common insertion sites ( CISs ) were identified using a multi-scale Gaussian Kernel Convolution approach [24] yielding 771 significant CISs compared to 0–3 expected from simulations of random integration ( Table S1 ) . A list of all RIS is provided as a . bed file for visualisation in genome browsers , version mm9 ( Table S2 ) . Notably , analysis of CIS accrual by number of tumours indicated that this system is approaching saturation and that virtually all the retrievable CISs have been detected ( Figure 1D ) . Target genes affected by integration at CISs were identified by computational methods [25] followed by manual curation . All 14 target genes identified by shotgun cloning methods [26] featured prominently ( Figure 1E; Table S3 ) . There was a positive correlation between the number of clones previously detected by shot-gun cloning and the number of 454 reads ( linear regression analysis; R = 0 . 56 ) showing that earlier lower powered methods detect only the “tip of the iceberg” of clonal expansion . While splinkerette/454 analysis is only semi-quantitative due to restriction enzyme site distribution and primary sequence constraints on PCR efficiency , we noted that the most abundant RIS corresponding to Pim-1 insertions were also detectable as rearrangements by Southern blot analysis ( Figure S2 ) . Moreover , the top 40 RISs ( by number of reads ) show few apparent passenger insertions , defined as isolated RIS far from any known target gene ( 5/40 ) , although these predominate ( 85% ) in the total population of 12 , 485 RISs . The possibility that most of these clones have acquired two separate driver insertions without any passenger insertions appears unlikely , suggesting that most highly proliferative clones contained only a single provirus . If the progression network consists of target genes that can complete the oncogenic transformation process , they would be expected to feature strongly in the dominant clones found in end-stage MoMLV-induced lymphomas . To test this assumption , we examined the overlap between the 771 progression CISs in this study with a meta-analysis by Kool and co-workers involving CISs identified by shotgun cloning of 19 , 923 unique RIS from 977 MoMLV-induced lymphomas of wild-type or tumour suppressor deficient mice [3] . Due to the lower sensitivity of the approach , these CISs should be enriched for major expanded tumour clones . A remarkable 346 CISs ( 45% ) were found in common between the Kool CISs and the progression CISs , indicating that a significant proportion of the target genes involved have been implicated previously as drivers of lymphoma development ( Figure 2A , Table S4 ) . Preferential integration of γ-retroviruses around transcriptional initiation sites is an established phenomenon [19] and on the basis of this and further evidence of non-random behaviour it has been argued that the observation of a CIS is insufficient evidence that post-integration selection for growth has occurred , particularly in large scale analyses [27] . While the ideal comparison with the progression CISs identified here would be normal thymocytes immediately after infection , there are significant technical challenges in obtaining a reliable in vivo baseline measurement due to the kinetics of infection and ongoing replication . We therefore chose to compare some aspects of our data to a published large-scale study of human CD34+ cells obtained after in vitro infection with a non-replicating MLV vector . This study by Cattoglio et al . is described as ‘near-baseline’ , as analysis was not carried out until 10 days post-transduction [21] . Notably , preference for transcriptional start sites was relaxed in the CISs observed in our study and this trend was more evident still in CISs with an orientation bias , consistent with the increasing importance of post-integration oncogenic selection in this subset ( Figure 2B ) . Moreover , we noted that most of the highly targeted CISs displayed the pronounced orientation bias that is classically associated with enhancer-mode gene activation [1] . As orientation bias does not arise at the level of integration [28] , this feature provides direct evidence of post-integration clonal selection . Stringent filtering of CISs for orientation bias yielded 17 examples which we will refer to as biased CISs ( Figure 2C; Table S5 ) . We applied the same approach to the Cattoglio ‘near-baseline’ dataset [21] and found no clusters with significant orientation bias after correction for multiple testing . CIS target genes displaying strong orientation bias were also the most frequently targeted and often displayed the greatest levels of clonal expansion , suggesting that enhancer mode activation is the most efficient process by which MoMLV drives lymphoma progression . An interesting outcome of this analysis shown in Figure 2D is that it provides strong support for the Myb gene as the target of long-range activation by insertions both 5′ and 3′ , including the CIS annotated as Ahi1 , in accord with hypotheses based on gene expression studies in lymphoma cell lines [29] , [30] . Further examples of genes subject to enhancer mode insertions are shown in Figures 2D and S3 . Evidence that the biased CIS targets form part of a larger progression network under selection was provided by KEGG pathway analysis which showed that some of the most frequent CIS targets ( e . g . Ccnd3 , Ccr7 , Pik3cd , Pik3r5 , Rasgrp1 ) map to metanodes that include many of the less frequent targets ( Figure S4 ) . Furthermore , KEGG pathway enrichment analysis showed that statistically significant over-representation of specific signalling pathways ( T-cell receptor , chemokine , JAK-STAT ) was evident even when the top 50–100 target genes were excluded from the analysis ( P = <1×10E-5 ) , arguing that oncogenic selection may also be occurring at sites that harbour only a few insertions ( Figure S5 ) . While orientation bias is useful to identify oncogenic selection on a background of preferential integration , we noted that there was a second frequent CIS group defined by intragenic insertions that displayed no statistical bias in orientation . Evidence that these are also under oncogenic selection is provided by the fact that 17 of the 20 most frequent targets have been observed in end-stage lymphomas ( Table S6 ) and by the fact that a significant subset have annotation suggestive of tumour suppressor or oncogene function ( Ikzf3 , Mad1l1 , Als2 , Ppp1r16b , Prex1 , Ttc28 and Ptprc ) . The typical pattern of insertions distributed across the target genes is suggestive of a tumour suppressor role , although a role for oncogenic truncated isoforms is also plausible [1] , [9] , [31] Although the majority of top ranking MLV target genes were shared between our progression dataset and the Kool meta-analysis of end-stage lymphomas , there were also notable differences . This was evident from comparison of CIS peak heights and relative rank order of CISs between the datasets where the most discordant examples are listed ( Table S7 ) . Oncogenic complementation was evident , with greatly reduced targeting of Myc/Pvt1 , Mycn and Runx family genes in the progression set . However , there was also a marked loss of selection for some major targets recorded by Kool et al . including Gfi1 and Notch1 . It appears that the combination of MYC and Runx2 in this context also renders these insertions redundant , which is intriguing as insertions at Gfi1 have been shown to be positively selected in some CD2-Runx2 lymphomas [18] . Also of interest was the large number of novel CISs in the progression set ( Table S8 , examples shown in Figure S3 ) . The most frequently targeted CIS targets displaying strong evidence of enhancer mode activation included Otx2 , a homeobox transcription factor which plays a major oncogenic role in medulloblastoma [32] but has not previously been observed in haematopoietic cancers and Myo16 , an atypical nuclear myosin with links to survival , cell cycle progression and PI3K signalling [33] . Moreover , a number of prominent targets for potentially disruptive intragenic insertions were unique to the progression set . These included Endou ( Pp11 ) , a placental poly-U endonuclease over-expressed in ovarian adenocarcinomas [34] , Xrra1 , which has been shown to modulate the response to X-ray irradiation [35] , and Ttc28 ( Tprbk ) , encoding a large tetratricopeptide domain protein that is regulated by p53 , complexes with BRCA1 and suppresses the growth of Ras-transformed cells [36] . A previously published analysis of preferential integration targets in early passage CD34+ cells showed a good correlation between basal transcriptional levels and integration frequency [21] . To test whether progression RIS targets were also selected by their high transcription rates in premalignant cells , we compared the transcriptomes of Runx2/MYC and control thymus at 10 days of age , several weeks before clonal tumours emerge . Figure 3 shows expression scatter plots for all gene probes . Basal expression of the most prominent progression targets was widely variable , and only Ccnd1 showed significant up-regulation compared to control thymus . Moreover , the frequent MoMLV targets that were not enriched in the progression network showed a similarly wide distribution with regard to expression levels . The exquisite selection by RIM of specific members of multigene families ( e . g . Jdp2 , D cyclins ) also appeared to be poorly correlated with expression level , strengthening evidence for post-integration selection as the predominant force shaping the progression network . Frequent targeting of Ccr7 , and to a lesser extent Ccr9 , is interesting in view of their central roles in mediating T-cell progenitor homing to thymus [37] , [38] . Moreover , Ccr7 has been reported as a mediator of progression and homing to lymph nodes in multiple tumour types , and to stimulate survival pathways by autocrine or paracrine mechanisms [39] . The cognate ligands for Ccr7 and Ccr9 ( Ccl21a , Ccl19 , Ccl25 ) , are highly expressed in normal thymus , but intriguingly were significantly down-regulated genes in premalignant organs ( validation shown in Figure S6 ) . The respective chemokine genes are normally expressed only in non-lymphoid elements of the thymus including epithelial cells [40] . The possibility that these genes were aberrantly activated to drive autocrine growth in the lymphoma cells was tested by direct analysis of isolated lymphoma cells ( Figure S7 ) . However , expression of the ligand genes was below detectable levels in Runx2/MYC or CD2-MYC/p53 null lymphoma cells suggesting that activation of Ccr7/9 provides a growth advantage by a paracrine mechanism that is dependent on thymic stroma . Falling expression of ligand genes in 10-day Runx2/MYC thymus may be due to down-regulation or simple occlusion of non-lymphoid cells by nascent lymphoma cells , which is virtually complete at later stages ( Figure S1 ) . To compare the progression network with genes selected during earlier events in tumorigenesis , a second RIM/DS barcode screen was conducted , including MoMLV-infected end-stage lymphomas from parental CD2-MYC , -Runx2 and wild-type mice with a subset of Runx2/MYC progressing tumours ( Figure 4 , Table S9 ) . All insertions , sorted by genotype are provided as a . bed file for visualisation in genome browsers , version mm9 ( Table S10 ) . MYC and Runx2 transgenes each cooperate with MoMLV to accelerate lymphoma onset to around 60 days post-infection [8] , [17] , [18] . Compared to MYC/Runx2 , the other three tumour sets yielded many more reads , but from a much smaller number of unique RISs , reflecting the presence of highly expanded tumour clones ( Figure 4A , B ) . The massive number of RISs per tumour ( 221–276 ) shows that in MoMLV lymphomas the predominant end-stage clones co-exist with a polyclonal background of minor populations . Application of an abundance threshold of 100 copies ( Figure 4B ) yielded a RIS number close to that expected from Southern blot analyses of end-stage MoMLV tumours that estimated 4–6 RISs in each dominant clone [41] . In most cases this cut-off correlated well with previous direct analyses for gene rearrangement [18] , [42] , although rearrangements of Myc detected by Southern blot in two of the CD2-Runx2 tumours analysed here failed to register in the splinkerette/454 analysis . Occasionally ‘missing’ clones might be explained by technical limitation e . g . due to sequence drift in primer sequences . In this regard , it is noted that the bias towards Myc family insertions was less marked here than in a Southern blot-based analysis of a larger CD2-Runx2 tumour cohort [18] . Nevertheless , there were clear and profound differences between cohorts , as MYC transgenic tumours resolved into fewer clones with substantially greater clonal enrichment compared to other genotypes while the double transgenics showed greater complexity as expected ( Fig . 4B , C ) . This apparent difference in the mode of tumour acceleration is interesting as CD2-Runx2 mice harbour an expanded population of transformation-prone thymocytes , which has no parallel in CD2-MYC mice , most of which remain healthy with no obvious abnormality [8] , [16] , [43] . The most striking features of the single transgenic tumours were evident when the most abundant RISs were sorted according to gene family ( Figure 4C , D ) . High copy RIS mapping to Runx2 or Runx3 were almost ubiquitous in , but exclusive to , CD2-MYC tumours ( P = 0 . 0001 , Fisher's Exact Test ) . A number of high abundance RIS mapped far upstream of Runx2 , adding this gene to the list of those subject to long-range activation . Only two tumours displayed no detectable Runx insertion . Another salient observation from analysis of the end-stage lymphomas was that the low abundance RIS left after subtraction of the major clones frequently correspond to progression network genes ( Tables S11 , S12 ) . It is conceivable that these represent tumour subclones that have acquired a further hit of proviral insertion , although the alternative possibility that these represent insertions in prelymphoma cells cannot be excluded . The possibility that this background reflects preferential integration in untransformed cells appears unlikely , as such cells form only a tiny fraction of the thymic mass and the hallmark orientation bias at major targets ( Figure 1C ) is also evident in these minor populations . Moreover , expanded RIS indicative of third hit genes in CD2-MYC , CD2-Runx2 and wild-type mice appeared to be selected from a broad cross-section of the progression network , with the ‘winners’ of the progression race largely recapitulating the expansion rate measured by earlier analysis of the progression network ( Figure 1D ) . We reasoned that specific ‘second hit’ collaborating genes would be distinguishable from progression genes on the basis of ( a ) positive selection in lymphomas of single transgenic mice compared to wild-type and ( b ) loss of selection or reduction to background levels in double transgenics . As expected , the Runx genes ( Runx2 and Runx3 ) and Myc family targets ( Myc , Mycn ) conformed to this pattern , being selected in CD2-MYC and CD2-Runx2 respectively and effectively disappearing from the double transgenic tumours ( Figure 4C ) . Surprisingly , inspection of the entire CIS list revealed only one other target gene with statistically significant correspondence to this pattern: intragenic insertions in Ikzf1 were significantly more abundant in CD2-Runx2 transgenic tumours than in the other three genotypes and showed more frequent representation in dominant clones ( Figure 5 ) . Intriguingly , analysis of the CD34+ ‘random integration’ vector dataset [21] shows two hotspots for integration in the human IKZF1 gene that correspond to active chromatin marks . The murine Ikzf1 gene showed a similar background pattern , although 3–4 clusters of insertions could be discerned in the murine gene . These observations suggest a two-step model for targeting of Ikzf1 by MoMLV , with preferential integration at sensitive sites within the gene leading to sustained clonal expansion only in the presence of a collaborating lesion such as deregulated Runx expression ( Figure 5B ) . Table S13 summarises the genes showing strongest evidence of complementation in parental transgenic mice . In addition , there is evidence of reduced selection for Gfi1 and Notch1 insertions on the CD2-MYC background which directly mirrors findings on the progression set compared to common MoMLV targets ( Table S7 ) suggesting that this bias is conferred by the CD2-MYC transgene . Targeting of both genes in wild-type controls and CD2-Runx2 in this study rules out mouse strain differences as the basis of this phenomenon . Notably , Notch1 has been shown to block p53-dependent apoptosis due to Myc over-expression [44] , while Gfi1 has recently been shown to modulate p53 responses indirectly by altering protein methylation [45] . The latter finding illuminates early RIM screens of Eμ-Myc mice which suggested that Gfi1 and Bmi1 belong to the same complementation group [7] , and Bmi1 is known to control p53 responses by transcriptional suppression of Arf [46] . As we have shown that Runx2 also inhibits Myc-induced apoptosis in vivo and that the Runx2/MYC combination neutralises selection for loss of p53 [16] , we propose the model in Figure 6 to account for the respective gene interactions in different transgenic backgrounds in a three-hit model of MoMLV lymphomagenesis . The extent to which the pathways targeted in retrovirus-induced lymphomas overlap with similar tumours of non-viral origin , including human cancers , is also of considerable interest . We compared the comprehensive CIS database with regions of chromosomal gain and loss described in a previous study of T-cell lymphomas arising in mice defective in telomerase , p53 and ATM ( ‘TKO’ ) mice [47] , where a strong overlap was noted with human T-ALL . Remarkably , 16/18 regions of syntenic deletion or amplification contained CISs , corresponding to 43/771 CISs ( for this overlap P = <0 . 0001; Table S14 ) . Notably , no known cancer genes could be found at the majority of these domains [47] , suggesting that the genes targeted at these CISs represent candidates for gain or loss of function that is conserved between human and mouse cancers . Significantly , many of the target genes display intragenic insertions , particularly for the deleted regions ( 13/22 ) . An interesting example is Gpr132 , located on chromosome 12 , which encodes a G-protein coupled receptor with apparent tumour suppressor activity [48] .
In this study we examined an established system of oncogene cooperation and retroviral acceleration using a deep sequencing ( DS ) platform . RIM/DS increases sensitivity of RIS detection by almost two orders of magnitude over earlier methodologies [9] and when applied to a lymphoma progression model shows evidence of saturation , indicating that all relevant major CISs have been obtained . The remarkable observation that much of the large repertoire of MoMLV target genes from almost one thousand end-stage T-cell lymphomas can be found in the progression network from only 28 lymphomas shows the enormous potential of RIM/DS when applied to polyclonal populations under strong selection . While statistical and pathway analyses provide useful tools to discriminate genes under oncogenic selection from preferential integration , our findings suggest that the phenomena may not be completely separable . The example of Ikzf1 illustrates the principle whereby a gene may be selectively targeted by γ-retroviral integration but leads to clonal expansion in the presence of a complementary oncogenic programme provided in this case by Runx gene activation . It has been demonstrated recently that γ-retroviral integration at transcriptional start sites is a consequence of interaction with BET chromatin regulators that direct the process towards genomic regions rich in acetylated histones [49] , [50] . The integration specificity of γ-retroviruses is clearly fundamental to their efficient replication and transmission in nature . In wild-type mice , the rate of oncogenic transformation due to successive integration events is reduced by retroviral interference , but the process is accelerated in oncogene transgenic mice where fewer hits are required . The implications of our analyses are also interesting for retroviral vector-based gene therapy . As the most potently selected insertions mediate enhancer-mode gene activation , the removal of enhancer elements in self-inactivating vectors [51] is likely to improve safety margins . However , failure to deal with the targeting apparatus will leave a residual risk , particularly for gene disruption events which , from their lack of obvious orientation bias , may not require strong enhancer function ( e . g . at Ikzf1 ) . While intrinsic preference for integration at transcriptional start sites and other chromatin features [19]–[21] creates the platform on which oncogenic selection operates , it is clear that post-integration selection events play a decisive role in shaping the genetic profile of end-stage tumours . The progression network is highly adapted to the T-cell environment but is not simply a cross-section of highly expressed and therefore available target genes . This principle is illustrated by the strong selection for specific members of multigene families ( e . g . Jdp2 , D cyclins ) that show no correlation with basal transcription levels . Similarly the targeting of novel genes that were not seen in previous large-scale screens of MoMLV-induced T-cell lymphomas ( e . g . Otx2 , Myo16 ) is not merely due to their up-regulation in the background of the Runx2/MYC model . These findings suggest that it will be of value to employ RIM/DS to probe the growth checkpoint networks in tissues and cell lineages that have been less well explored to date . While most of the functionally annotated progression network genes are predicted to confer autonomous proliferation , an exception to this rule was provided by the frequent activation of Ccr7 and Ccr9 , which in their normal developmental roles promote T-cell homing to thymus and ligand-dependent survival and proliferation [37] . Moreover , Ccr7 is stimulated by Notch signalling [52] , and we would predict that retroviral activation bypasses this requirement . It appears that the result of Ccr7/9 activation in Runx2/MYC lymphomas is likely to be paracrine growth stimulation , as expression of the cognate ligands ( Ccl19 , 21 , 25 ) is restricted to thymic stromal cells . Moreover , declining levels of ligand transcripts in Runx2/MYC thymus offers a rationale for the accelerated dissemination of lymphoma cells towards highly expressing peripheral lymphoid tissues [9] . Export of lymphoma cells with Ccr7 insertions is also in accord with the relatively low read/RIS ratio in primary thymic lymphomas . Identification of Ccr7 as a major target highlights the complementary value of RIM screening , as this gene does not appear to be subject to mutation or amplification in human cancer , yet is required for CNS metastasis of human leukaemia cells [52] . Comparison of the progression network with a large scale meta-analysis of MoMLV targets in T-cell lymphomas from various genetic backgrounds [3] showed that the principles of complementation apply where the two germ-line oncogenes are present , as insertions at Myc and Runx family members were massively under-represented in the progression set . Moreover , while most major targets overlapped strongly , a few prominent targets including Gfi1 and Notch1 were also greatly diminished in the progression network . Our second RIM/DS of parental transgenic mice shed further light on this observation , as the CD2-MYC parental transgenic system in particular did not select for these targets but instead showed virtual dependence on activation of a Runx family gene with the order Runx2>Runx3>Runx1 in targeting frequency in accord with previous observations [53]–[55] . Comparison of several Myc transgenic model systems ( CD2-MYC , Eμ-Myc , Mmtv ( d ) -Myc ) shows that these have massively divergent preferences for collaborating genes detected by RIM , presumably reflecting the lineage and stage-specificity of Myc expression control [7] , [42] , [56] . However , it is notable that all of these potently selected collaborating genes share the ability to suppress the p53 response in the context of activated Myc [16] , [44] , [46] , [57] . There is an obvious parallel with the observation that the combination of CD2-Runx2/MYC overcomes the requirement for genetic inactivation of the p53 pathway [16] , providing a rationale for the reduced selection for Notch and Gfi1 on this background . The foregoing observations invite the model presented in Figure 6 , where the interaction of this small gene set is presented as a bottleneck to transformation in contrast to the broad range of progression genes that can be recruited at later stages . In addition to the simple outline shown here , it appears that the MoMLV ‘core’ gene programme can also neutralise p53-independent failsafe pathways , as p53 deficiency has relatively modest effects on MoMLV-induced tumour onset and target gene spectrum [5] , [14] , [15] . It should also be noted that at least some of the genes in the progression network can also serve as initiators when expressed as transgenes , showing that the mutational order may not be fixed [58]–[60] . Why do the major collaborating gene targets vary so markedly between Myc transgenic models ? The most obvious rationale is presented by the lineage and stage-specificity of Myc expression . RIM targeting of Bmi1 is largely a feature of B-cell lymphomas in the mouse [7] , while Notch targeting predominates in the CD4+CD8+ lymphomas of Mmtv ( d ) -Myc mice [56] . The CD2 LCR confers strong T-cell specificity but is also active in B-cells [61] , implying that its developmental activation may occur at the level of committed lymphoid progenitors . High level Myc expression in this niche appears to lead to cell death , unless combined with loss of p53 or an activated Runx allele [17] , [62] , [63] . We hypothesise that Notch1 or Gfi1 pathways are not available for RIM targeting at this stage and that Runx2 , the ‘bone-specific’ family member , which is also transcriptionally active in early haematopoietic development [64] , becomes the primary target for activation in this niche . As mounting evidence indicates that Runx family members are downstream of Notch signalling in expression control and effector functions [65] , it is tempting to suggest that dual activation of Runx and Myc supplants the need for activation of Notch . The model we propose has implications for therapeutic targeting of Notch signalling with γ-secretase inhibitors [66] , as up-regulation of Runx and Myc may represent another pathway to resistance . Although CD2-Runx2 selects strongly for activation of Myc family genes by RIM [18] it appears less critically dependent , possibly due to the survival of Runx2 expressing thymocytes as a premalignant , slowly proliferating population blocked at the DN/CD8ISP stage [43] . This study shows that Ikzf1 is also favoured as a collaborating target on this background . Notably , Ikzf1 is a haplo-insufficient tumour suppressor that has been reported to act as a transcriptional suppressor of Myc [67] , while intragenic retroviral insertions lead to expression of truncated isoforms with dominant negative potential [31] . We therefore suggest that de-repression of Myc may be one of the consequences of Ikzf1 targeting that leads to its co-selection with Runx2 . It would interesting in this regard to test whether lymphomas of Runx2 transgenic mice with reduced Ikzf1 function [68] would show reduced RIM targeting of both Myc family genes and Ikzf1 . This analysis has wider implications for the genetics of human lymphomas and other cancers . It appears that the final step in lymphoid transformation by MLV can be accomplished by a wide range of genes with the common functional end-point of growth factor-independent proliferation . As the progression network also includes numerous genes that are mutated , amplified or deleted in human cancer ( Table S14 ) , it is tempting to suggest that many of the acquired mutations in human cancer are also late embellishments . Another important insight is provided by the evidence of a small network of genes ( Myc , Runx , Ikzf1 , Gfi1 , Notch1 , and Bmi1 ) that act in pairwise combinations to confer lymphoma self-renewal and overcome failsafe responses via the p53 pathway . It seems likely that this network operates under normal physiological conditions to licence cell growth and is co-ordinately subverted in cells carrying mutations in the pathways . The recent description of Gfi1 as an ‘oncorequisite’ factor that is rarely directly mutated but nevertheless required for growth of ALL cells [45] highlights the potential for targeting this network . The Runx genes are heavily implicated in human leukaemia but show paradoxical features of either gain or loss of function in disease subsets [69] . The demonstration here that Runx activation is virtually essential for MYC transformation of early murine T-cell lymphoma suggests that it may be fruitful to examine the requirement for RUNX function in human leukaemia/lymphomas driven by amplified MYC or NOTCH/IKZF1 mutations .
Animals were routinely monitored and sacrificed when showing signs of ill health in line with the UK Animals ( Scientific Procedures ) Act , 1986 . CD2-MYC , CD2-Runx2 , and CD2-MYC/CD2-Runx2 transgenic animals and maintenance were described previously [9] . Neonates were infected within 24 hours of birth with ∼105 infectious units of MoMLV as previously described [42] . Littermate-matched genotype controls were used to control for mouse strain . DNA was extracted from approximately 20 mg of frozen enlarged lymphoid/tumour tissue using Gentra Puregene Genomic DNA Purification Kit ( Qiagen , UK ) according to the manufacturer's instructions . Isolation of the retroviral insertion sites from the tissues was performed using splinkerette PCR to produce barcoded PCR products that were pooled and sequenced on 454 GS-FLX sequencers ( Roche Diagnostics platform ) as described previously [70] , [71] . The restriction enzymes used to digest the genomic DNA were Sau3AI and Tsp509I , and the enzyme used to digest MoMLV DNA was EcoRV . Processing of 454 reads , identification of insertion sites , and Gaussian kernel convolution ( GKC ) statistical methods used to identify common insertion sites ( CISs ) have been described previously [6] , [24] , [71] , [72] . In summary , 454 reads were mapped to the mouse mm9 genome assembly , where the only modification to the previous alignment procedure was the removal of the stringency check as to whether an alignment was located neighbouring a TA dinucleotide site ( the insertion locations preferred by Sleeping Beauty transposons on which the bioinformatics processing method was developed ) . Reads from the same sample whose start genomic locations aligned within three nucleotides of each other were merged together . Reads from the same sample that were more than three nucleotides apart were considered independent integration events . CISs were identified using the multi-scale GKC approach [6] , [24] In order to determine whether the MLV screen had reached some level of saturation , the Gaussian Kernel Convolution ( GKC ) CIS calls from all 28 samples were analysed using the ACT software package [73] . ACT considers genomic locations generated by multiple samples for specific biological phenomenon under study ( e . g . ChIP-seq peaks ) to determine the saturation of a screen . The program considers the various combinations in which samples can be added so that the increase in base pair coverage is a range of values based on all the samples . The results can be depicted as a series of boxplots showing the increase in base pair coverage , where the boxplot at each position n on the x-axis shows the coverage values of all combinations of n samples . Boxplots that approach a horizontal asymptote indicate that the coverage has reached saturation . For the GKC CISs generated by all 28 samples , the insertion sites that contributed to CISs were extracted , resulting in a set of 7 , 485 sites . The insertion sites were then selected per sample and pseudo-kernels of 7 . 5k nucleotides either side of each insertion were applied to mimic GKC kernels of 15k nucleotides . Overlapping kernels within each sample were merged into continuous genomic regions . These 28 modified insertions files were then analysed using ACT . For each combination of samples the median values , and 25th and 75th percentiles were plotted using ggplot2 [74] . As a control , the 28 samples were re-analysed where the same number of insertion sites per sample were selected at random across the mouse genome . The pseudo-15k nucleotide kernels were applied . While the analysis does not produce a clear-cut asymptote this is to be expected due to the type of data under consideration . ACT was designed to analyse such data as ChIP-seq arrays for predicting transcription factor binding sites . In these scenarios ChIP-seq replicates should ideally report the same key binding sites/genomic locations . Hence across multiple samples the same locations should be reported . For MLV screens however , while insertions in the same gene will be found from different samples , the locations of the insertion sites will not overlap perfectly , even with the addition of the 15k nucleotide pseudo kernels . Hence each sample will introduce novel regions , such that the overall coverage will continue to increase even if the screen has truly reached a ‘saturation’ point . Also not all samples will contribute to all CISs . Different combinations of samples will thereby result in varying coverages , causing the coverage profile not to asymptote perfectly . The genomic coordinates of the ‘UCSC Genes’ set was downloaded via the UCSC genome browser for mouse assembly mm9 . Each of the 12 , 485 MoMLV integration sites was then mapped relative to the transcription start site ( TSS ) of its closest UCSC-defined ‘known’ gene . The Kool set of 19 , 923 mouse retroviral insertions sites was downloaded from the Mutapedia website ( http://mutapedia . nki . nl/ ) [3] . In the original paper , 596 CISs were identified using the GKC statistical framework with a fixed kernel width of 30k nucleotides . The insertion sites were re-analysed using the same multi-scale kernel approach that was applied to the MoMLV insertion sites . As a result of the multi-scale kernels and a less stringent cut-off value , 977 CISs were identified . Defining the width of a CIS as spanning the minimum and maximum genomic coordinates of insertion sites that contribute to a CIS , CISs were compared between the progression set and the re-analysed Kool set for overlaps . CISs were called overlapping if at least one nucleotide was overlapping between the two CIS sets . A set of 32 , 592 human MLV-based vector integration sites was kindly provided by Cattoglio and co-workers as previously published [21] . In the original study genomic regions were considered as significant if three or more integration sites were found clustered within regions of 12 , 587 nucleotides . This threshold was applied to the 32 , 592 integrations sites resulting in the identification of 3 , 453 clusters . Taking the integration sites within the clusters , a similar Fisher's exact test method was used to assess the orientation bias of the integration sites as for the MoMLV CISs . Following multiple test correction no clusters exhibited any orientation bias . RNA was isolated and purified from the thymuses of 10 day old wild type and CD2-MYC/Runx2 double transgenic mice using an RNeasy Mini Kit as per the manufacturer's instructions ( Qiagen , UK ) with mechanical lysis using a pellet pestle in a microfuge tube ( Sigma ) . RNA purity was assessed using a Nanodrop 2000 Spectrophotometer ( Thermo Scientific ) , and integrity verified using the Agilent 2100 Bioanalyser with RNA 6000 Nano Reagents kit ( Agilent Biotechnologies ) as per the manufacturer's protocol . Whole genome expression profiling was performed using Affymetrix mouse GeneChip microarrays ( MoGene-1 ) in triplicate as per the manufacturer's protocol ( Affymetrix , UK ) . Data analysis was carried out using the Partek Genomic Suite ( Partek Inc . , St . Louis , MO , USA ) . Briefly , after Robust Multichip Average normalisation [76] with GC content pre-background adjustment , the differentially expression analysis was performed using ANOVA . Multiple testing correction was done using the ‘q value’ cut-off [77] with gene changes of p<0 . 05 considered significant . Graphical representations of data were prepared using CLC Genomics Workbench 4 .
|
Cancers are known to arise by a series of mutational and non-mutational ( epigenetic ) events but the advent of cancer genome sequencing highlights the growing challenge of separating important ( driver ) from irrelevant ( passenger ) mutations . Retroviruses that induce cancer by inserting into host DNA and thereby altering key genes are valuable tools because they act as ‘tags’ to identify the critical targets . In this study we combined retroviral tagging with next generation sequencing to achieve a comprehensive description of lymphoma development and progression in transgenic mouse model systems . Our study suggests that three events may be sufficient for lymphoma development and identifies a genetic bottleneck at a small gene set that regulates tumour cell self-renewal , including the Myc oncogene and the p53 tumour suppressor . In contrast , many genes can provide the final step where the lymphoma cell acquires the ability to divide independently of external stimuli . As many of the target genes are conserved and play roles in cancers of non-viral origin , this study may provide a paradigm for the gene interactions that underlie cancer biology . It also elucidates the risks entailed in the recent use of retrovirus-based vectors for human gene therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"animal",
"models",
"sequence",
"analysis",
"medicine",
"oncology",
"model",
"organisms",
"biology",
"genomics",
"mouse",
"computational",
"biology"
] |
2014
|
Insertional Mutagenesis and Deep Profiling Reveals Gene Hierarchies and a Myc/p53-Dependent Bottleneck in Lymphomagenesis
|
Human alveolar echinococcosis ( AE ) is a severe zoonotic disease caused by the metacestode stage of Echinococcus multilocularis . AE is commonly associated with a long incubation period that may last for more than ten years . The objective of this systematic literature review was to identify and summarize the current knowledge on statistically relevant potential risk factors ( PRFs ) associated with AE in humans . Six bibliographic databases were searched , generating a total of 1 , 009 publications . Following the removal of duplicate records and the exclusion of papers that failed to meet the criteria of a previously agreed a priori protocol , 23 publications were retained; however , 6 of these did not contain data in a format that allowed their inclusion in the meta-analysis . The remaining 17 publications ( 6 case-control and 11 cross-sectional studies ) were meta-analysed to investigate associations between AE and PRFs . Pooled odds ratios ( OR ) were used as a measure of effect and separately analysed for case-control and cross-sectional studies . In the case-control studies , the following PRFs for human AE showed higher odds of outcome: “dog ownership” , “cat ownership” , “have a kitchen garden” , “occupation: farmer” , “haymaking in meadows not adjacent to water” , “went to forests for vocational reasons” , “chewed grass” and “hunting / handling foxes” . In the cross-sectional studies , the following PRFs showed higher odds of outcome: “dog ownership” , “play with dogs” , “gender: female” , “age over 20 years” , “ethnic group: Tibetan” , “low income” , “source of drinking water other than well or tap” , “occupation: herding” and “low education” . Our meta-analysis confirmed that the chance of AE transmission through ingestion of food and water contaminated with E . multilocularis eggs exists , but showed also that food- and water-borne PRFs do not significantly increase the risk of infection . This systematic review analysed international peer-reviewed articles that have over the years contributed to our current understanding of the epidemiology of human AE . The identification of potential risk factors may help researchers and decision makers improve surveillance and/or preventive measures that aim at decreasing human infection with E . multilocularis . More primary studies are needed to confirm potential risk factors and their role in the epidemiology of human AE .
Alveolar echinococcosis ( AE ) is considered as one of the most dangerous parasitic zoonoses occurring in the northern hemisphere . The life cycle of the parasite is maintained by definitive hosts shedding eggs with the faeces and the subsequent ingestion of these eggs by suitable intermediate hosts . Definitive hosts become infected by eating intermediate hosts , in which the metacestode stage develops . Carnivores , in particular the red fox in Europe and the raccoon dog among wildlife , as well as the dog and to a lesser extent the cat , represent definitive hosts for the parasite [1–4] . In other epidemiological settings , for example in the People's Republic of China , dogs may , in addition to foxes , play an important role in the transmission of AE [5] . Small rodents , mainly voles , serve as natural intermediate hosts . Humans are aberrant intermediate hosts and are affected by the metacestode stage of the parasite . They become infected through the ingestion of E . multilocularis eggs . After an initial asymptomatic incubation period of many years , humans may develop AE , which can be lethal if left untreated [6] . Diagnosis of probable AE is based on imaging techniques such as abdominal ultrasound , magnetic resonance , and computer tomography as well as detection of specific serum antibodies [7] . However , case confirmation requires histopathology or positive PCR on a fine-needle biopsy , which is particularly important in areas co-endemic for AE and cystic echinococcosis ( CE ) . Since AE is a rare disease with a long incubation period of several years , identification and evaluation of potential risk factors ( PRFs ) associated with infection is difficult . Moreover , the duration of the asymptomatic or paucisymptomatic period and the associated nonspecific symptoms in humans contribute to the massive underreporting of AE . Although in some regions considerable attention is afforded to the disease by both the general public and the media , AE is not yet mandatorily notifiable in most affected countries [8] . Human infection depends on behavioural and socio-economic variables that favour close contact with E . multilocularis eggs and their inadvertent ingestion [9 , 10] . Additionally , risk factors and the geographic distribution of AE may differ from country to country and more importantly between regions as the infection is influenced by both biotic and abiotic factors . The objective of this study was to summarize evidence relating to both proven and potential risk factors associated with E . multilocularis human infection by conducting a systematic review ( SR ) and meta-analysis of studies in accordance with the Cochrane and PRISMA group guidelines . The results of this SR may help establish or improve existing surveillance systems to prevent human infection with E . multilocularis .
This SR was essentially performed using the same methodology as in recently published SR on PRFs associated with human cystic echinococcosis [11] . Literature databases were searched using keywords associated with the Boolean operators AND/OR , the question mark ( ? ) and the hash mark ( # ) . The question mark ( ? ) expanded the search by looking for words with similar prefixes using more than one letter ( i . e . “echinococc ? ” was used to search for “echinococcus” , “echinococci” , “echinococcosis” and “echinococcoses” ) , whereas the hash mark ( “#” ) expanded the search by looking for words with similar prefixes using one letter ( i . e . dog# was used to search for “dog” or “dogs” ) . The strategy developed in PubMed/Medline used queries for papers reporting abstracts on risk factors related to human AE . Thus , the final strings used for the search were [echinococcus multilocularis OR ( echinococcus AND multilocularis ) OR e# multilocularis OR alveolar echinococcosis OR a# echinococcosis] AND ( risk factor# OR risk# ) AND ( human# OR people OR person OR man OR men OR women OR woman OR patient# ) . The following inclusion criteria were used: primary research studies published from 1950 onwards or in press at the time of the search; articles published in English , German , French , Polish , Finnish , Dutch , Spanish or Italian; studies with a case-control , cross-sectional or cohort design . Articles were excluded if they contained only the results of descriptive studies ( including case reports ) , if a data-driven assessment of PRFs for human infection with E . multilocularis was missing , if agents other than E . multilocularis ( e . g . Echinococcus granulosus sensu lato ) were investigated or if no difference was made between AE and CE . In addition , review articles , letters or editorials without original data , duplicated data and articles with full texts written in languages other than those mentioned above were not included . This SR and meta-analyses followed the Cochrane and PRISMA ( Preferred Reporting Items for SRs and Meta-analyses ) Group guidelines [12] . The PRISMA 2009 checklist is provided as supporting information ( S1 Checklist ) . The platform used for searching databases was STN International–Fiz Karlsruhe ( http://www . fiz-karlsruhe . de/stn . html ? &L=1 ) . The first online electronic search was carried out on 5th November 2013 and updated on 11th February 2015 and again on 1st April 2016 to identify papers published since the initial search . The platform used to carry out the literature search included six bibliographic databases: Medical Literature Analysis and Retrieval System Online ( MEDLINE ) , Excerpta Medica Database ( EMBASE ) , Science Citation Index ( SciSearch ) , Biological Abstracts ( BIOSIS ) , Centre for Agricultural Bioscience International ( CABI ) and Google Scholar . Article selection was carried out using a three step process: first , duplicate articles were removed , then titles and abstracts were read taking the keywords into account , and finally full texts were screened for eligibility . Data were extracted from elegible studies using standardized Microsoft Excel tables ( Microsoft Office 2010 , version 14 . 0 ) . The following data were recorded in the extraction tables: paper identification ( ID , sub-ID , first author , year of publication , title , journal , volume , page numbers ) , brief study description , study design ( case-control , cross-sectional and cohort study ) , year of study , geographic area , country and diagnostic method ( ultrasound , surgery , molecular identification and serology ) , PRFs and quality assessment . Data were extracted by two independent researchers and any disagreements were resolved by consensus among the researchers using the standardized extraction forms to guarantee consistency and accuracy . Articles meeting the inclusion criteria were evaluated , and data relating to PRFs were extracted according to the following groups: association with dogs/cats , gender , age , familiar or ethnic clusters , living in a rural area , having a kitchen garden , occupation , food/water contact , hunting/ handling foxes and socio-cultural status . The terms for PRFs were used as referred to in the respective studies and were not merged , even if they appeared to be similar ( e . g . dog ownership; allowed dog into the house; allowed dog into bedroom; was licked by dog; playing with dogs , brushed the dog´s fur; etc . ) . Data from studies based only on serology were extracted but due to the low sensitivity/specificity of this diagnostic approach , these studies were meta-analysed only when ultrasound results were also reported [13 , 14] . All patient medical data analysed in this study were anonymised and extracted from publications , in which they were reported in an aggregated form as case or population counts . No individual patient data were used in this study . Approval from an ethical committee or institutional review board was not necessary for this research . The quality assessment of the studies included in this SR was performed by two independent reviewers using the Newcastle-Ottawa Scale ( NOS ) according to the Cochrane Handbook for Systematic Reviews [15 , 16] . If the assessments differed , they were discussed until a consensus had been reached . Studies were scored in three domains: ( i ) selection of study groups , ( ii ) their comparability and ( iii ) the ascertainment of either the exposure or outcome of interest for case-control or cohort studies , respectively . The "selection" category covers four properties: use of an adequate case definition , representativeness of the cases , selection of controls and the definition of controls . The "exposure" category consists of three properties: ascertainment of exposure , the use of the same method of ascertainment for cases and controls , and information on the non-response rate . A study could be awarded a maximum of one star for each item in the categories "selection" and "exposure" . A maximum of two stars could be given for "comparability" ( study controls for one factor/ for any additional factor ) , i . e . a maximum score of 4 ( “selection” ) , 3 ( “exposure” ) and 2 ( “comparability” ) stars could be allocated leading to a maximum score of 9 . Statistical analysis was performed using the software Review Manager 5 . 2 ( RevMan Version 5 . 2 . Copenhagen: The Nordic Cochrane Centre , The Cochrane Collaboration , 2014; http://ims . cochrane . org/revman ) . The analyses were stratified in relation to the different PRFs reported in the included studies . Pooled odds ratios ( OR ) were used as a measure of effect and separately analysed for case-control and cross-sectional studies . Meta-analysis was performed when at least two studies reported data on the same PRF . OR , with the respective 95% confidence intervals ( CI ) , were calculated for each potential risk or protective factor ( if the factor was covered in more than two studies ) , and visualised using forest plots . The Cochran’s Q test was performed to assess the degree of heterogeneity between studies . I2 statistic was used to describe the percentage of total variation across studies . If the p-value of the Q test was < 0 . 05 and I2 was >50% , heterogeneity was inferred and a random-effect model was used . If heterogeneity was not detected , a fixed-effect model was adopted . Publication bias was quantified by inspection of funnel plots and computation of Egger [17] and Begg [18] test probability values . A meta-regression analysis was conducted for each potential risk or protective factor reported in more than two studies using SPSS for Windows ( IBM Corp . , 2013 . IBM SPSS Statistics for Windows , Version 22 . 0 , Armonk , Ney York , USA ) . Year of publication , total population and quality scores were considered as variables . For each analysis , a linear regression model was built by stepwise backward elimination of variables . The results of the analyses are presented with their beta coefficients and p-values . The threshold for statistical significance was set at p<0 . 05 .
A total of 1 , 009 potentially relevant publications were identified . Following the removal of 515 duplicates generated between databases , the titles and abstracts of 494 papers were screened and a further 416 papers were excluded ( Fig 1 ) . The full texts of the remaining 78 publications were read , and 55 papers were excluded ( S1 Table ) based on the following: no risk factor was reported , the publication failed to be a primary study , no control group could be identified , no data on patients were reported , the publication was characterized as a case report or no distinction was made between E . multilocularis and E . granulosus . A total of 23 publications were found eligible for inclusion in this SR ( S2 Table ) , of which 17 were subjected to meta-analysis to determine associations between human AE infection and potential risk or protective factors . The remaining 6 publications were unsuitable for meta-analysis as no OR for individual risk factors could be extracted . Of the 17 publications , data were separately extracted from case-control ( n = 6 ) [19–24] and cross-sectional studies ( n = 11 ) [10 , 25–34] ( S2 Table; S1 Flow Diagram ) . No cohort studies were identified . The geographical areas relevant to the 17 studies included in this review were China ( n = 11 ) [10 , 24–31 , 33 , 34] , Central Europe ( n = 5 ) [19–22; 32] and North America ( n = 1 ) [23] . Papers used in the current SR were published between 1988 and 2013 . When studies ( i . e . individual publications ) were conducted on different groups of individuals ( for example pastoral versus urban communities [10] ) or in different areas ( for example regions at risk versus regions not at risk of AE [21] ) , they were divided into sub-studies and each sub-study was analysed separately . Eventually , 8 and 13 sub-studies were identified for case-control and cross-sectional studies , respectively and each of these two groups were meta-analysed separately . The results of the meta-analyses ( forest plots , OR and pooled OR with their respective 95% CI ) and potential publication bias ( funnel plots ) were separately recorded for case-control ( S1 Supplementary information ) and cross-sectional studies ( S2 Supplementary information ) . When the meta-analysis included only a small number of studies , it was not possible to assess publication bias using funnel plots . The assessment of the quality of the studies included in this SR was performed using NOS through the implementation of a ‘star system’ . Of the 6 case-control studies , 3 were allocated an 8-star rating [19 , 22 , 23] and the other 3 studies received a 9 [20] , 4 [24] and a 2-star score [21] . Within the 11 cross-sectional studies , 5 received a 8-star rating [10 , 26 , 28–30] , 3 a 6-star rating [25 , 33 , 34] , 2 a 7-star rating [31 , 32] and 1 a 9-star score [27] . Sixteen PRFs were identified from case-control studies and meta-analyses were performed on six papers corresponding to eight sub-studies . Most studies originated from endemic areas in Central Europe , namely Austria ( n = 1 ) [22] , France ( n = 1 ) [21] , Germany ( n = 1 ) [19] , France-Germany-Switzerland ( n = 1 ) [20]; North America ( Alaska ) ( n = 1 ) [23] and Asia ( People's Republic of China ) ( n = 1 ) [24] . All these retrospective studies were hospital-based and included control groups that were not affected by AE but had demographic characteristics similar to those of the AE patients . The following groups of PRFs were investigated in this meta-analysis ( PRFs are termed as reported in the publications ) : 4 pet-related ( “dog ownership” , “allowed dog into the house” , “play with dogs” , “cat ownership” ) , 4 food-related ( “ate mushrooms” , “consumption of wild vegetables and fruit” , “ate unwashed strawberries” , “having a kitchen garden” ) , 7 related to working or recreational activities in rural areas ( “hunting” , “hunting / handling foxes” , “occupation: farmer” , “haymaking in meadows not adjacent to water” , “went to forests for vocational reasons” , “chewed grass” , “live in a rural area” ) and one potentially protective genetic factor ( “human leukocyte antigen , HLA” ) . Eight PRFs were associated with statistically significant increased OR ( test for overall effect , p<0 . 05 ) : “dog ownership” ( OR 2 . 50; 95% CI 1 . 73–3 . 62; p<0 . 00001 ) , “cat ownership” ( OR 2 . 63; 95% CI 1 . 42–4 . 85; p<0 . 002 ) , “having a kitchen garden” ( OR 5 . 21; 95% CI 2 . 65–10 . 22; p<0 . 00001 ) , “occupation: farmer” ( OR 4 . 50; 95% CI 2 . 74–7 . 39; p<0 . 00001 ) , “haymaking in meadows not adjacent to water” ( OR 3 . 50; 95% CI 1 . 63–7 . 55; p = 0 . 001 ) , “went to forests for vocational reasons” ( OR 2 . 61; 95% CI 1 . 13–6 . 05; p<0 . 03 ) , “chewed grass” ( OR 3 . 20; 95% CI 1 . 65–6 . 20; p = 0 . 00006 ) and “handling foxes” ( OR 2 . 27; 95% CI 1 . 35–3 . 81; p = 0 . 002 ) . The PRF “HLA” was statistically significant ( p = 0 . 003 ) with an OR of 0 . 50 ( 95% CI: 0 . 32–0 . 80 ) , which indicated that particular human leucocyte antigens may be protective against AE infection . For seven PRFs , the results were not statistically significant: “allowed dog into the house” ( OR 1 . 80; 95% CI 0 . 90–3 . 62; p = 0 . 10 ) , “play with dogs” ( OR 1 . 24; 95% CI 0 . 39–3 . 91; p = 0 . 71 ) , “living in a rural area” ( OR 3 . 07; 95% CI 0 . 83–11 . 37; p = 0 . 09 ) , “ate unwashed strawberries” ( OR 1 . 39; 95% CI 0 . 87–2 . 23; p = 0 . 17 ) , “hunting” ( OR 1 . 13; 95% CI 0 . 69–1 . 83; p = 0 . 63 ) , “ate wild vegetables and fruit” ( OR 1 . 38; 95% CI 0 . 90–2 . 10; p = 0 . 14 ) and “ate mushrooms” ( OR 0 . 72; 95% CI 0 . 38–1 . 39; p = 0 . 33 ) . PRFs meta-analyzed for case-control studies are summarized in Table 1 . Forest plots , funnel plots and single weight of each publication contributing to the overall risk factors are presented in S1 supplementary information . Meta-analyses performed on eleven cross-sectional studies corresponding to thirteen sub-studies revealed thirteen PRFs . Most studies originated from endemic areas in the People's Republic of China , namely , Ningxia Hui autonomous region ( n = 2 ) [29 , 34] , Sichuan province ( n = 4 ) [10 , 25 , 26 , 33] , Gansu province ( n = 2 ) [30 , 31] , Qinghai province ( n = 1 ) [28] , Sichuan and Qinghai provinces ( n = 1 ) [27] and one from Europe ( Germany ) [32] . All studies were community-based ultrasonography surveys . PFRs were grouped as follows: two PFRs were dog-related ( “dog ownership” , “play with dogs” ) , 2 were linked to drinking water ( “source of drinking water other than well or tap” , “drinking non-boiled water” ) , 3 were related to working activities ( “occupation: farmer” , “occupation: herder” , “hunting/ handling foxes” ) , 4 were socio-culturally related ( “ethnic group: Tibetan” , “low income” , “low education” , “hand wash before eating” ) and two could not be attributed to a specific group ( “gender: female” , “age over 20 years” ) . “Low education” was differentiated into having no education or attended primary school versus secondary school , college or higher education . “Low income” was defined as having a salary of less than 5 . 000 Chinese Yuans ( RMB ) per year ( exchange rate 2005: 8 . 1 Yuan = 1 US $ ) . Nine PRFs were statistically significant and exposure was associated with increased OR ( test for overall effect , p<0 . 05 ) : “dog ownership” ( OR 1 . 95; 95% CI 1 . 52–2 . 51; p<0 . 00001 ) , “play with dogs” ( OR 3 . 48; 95% CI 2 . 20–5 . 52; p<0 . 00001 ) , “gender: female” ( OR 1 . 66; 95% CI 1 . 31–2 . 10; p<0 . 00001 ) , “age over 20 years” ( OR 3 . 65; 95% CI 1 . 15–11 . 62; p<0 . 03 ) , “ethnic group: Tibetan” ( OR 2 . 03; 95% CI 1 . 56–2 . 63; p<0 . 00001 ) , “low income” ( OR 3 . 92; 95% CI 2 . 42–6 . 36; p<0 . 00001 ) , “source of drinking water other than well or tap” ( OR 1 . 81; 95% CI 1 . 52–2 . 17; p<0 . 001 ) , “occupation: herder” ( OR 2 . 20; 95% CI 1 . 51–3 . 19; p<0 . 00001 ) and “low education” ( OR 4 . 81; 95% CI 2 . 73–8 . 48; p<0 . 00001 ) . The PRF “drinking non-boiled water” was protective ( OR<1 ) against infection at a statistically significant level ( OR 0 . 63; 95% CI 0 . 48–0 . 84; p = 0 . 002 ) . For three PRFs , it was not possible to determine their effect ( protective , risk or none ) : “hand washing before eating” ( OR 4 . 90; 95% CI 0 . 80–29 . 87; p = 0 . 08 ) , “occupation: farmer” ( OR 1 . 16; 95% CI 0 . 31–4 . 35; p = 0 . 82 ) and “hunting / handling foxes” ( OR 1 . 25; 95% CI 0 . 71–2 . 20; p = 0 . 44 ) . The PRFs meta-analyzed for cross-sectional studies are summarized in Table 2 . Forest plots , funnel plots and single weight of each publication contributing to the overall risk factors are presented in S2 supplementary information .
This SR and meta-analysis summarized the current evidence on PRFs associated with human AE infection . The initial search revealed more than 1 , 000 studies published during the last 65 years , a fact that emphasizes the broad relevance of this topic . The studies included in this SR refer to a total of 3 , 091 AE cases . The number of AE cases per study varies from one [32] to 577 [27] . The largest control group comprised 15 , 614 persons [27] . We cannot rule out , however , that some persons were enrolled in more than one study as all data were used in an anonymised form . This might have been the case in the study of Eiermann et al . ( 1998 ) , which included 64 patients from France and 34 patients from Germany [20] . The same patients may have also been enrolled in the studies of Piarroux ( 2013 ) [21] with 180 AE cases from France and Kern ( 2004 ) with 40 AE cases from Germany [19] . However , even if the same patients were included , the aim , approach and methods of the studies were different . While the study of Eiermann et al . ( 1998 ) [20] focused on immunological markers , the two other studies [19 , 21] analysed environmental risk factors . So , inclusion of identical people in different studies investigating different PRFs does not bias the results of our meta-analysis . During the SR process , papers were evaluated by quality assessment tools , but similar to previous SR , e . g . [35] , it was not possible to evaluate the validity of diagnostic techniques or to differentiate between “possible” , “probable” or “confirmed” diagnosis . In general , dog-related PRFs for human AE are common worldwide . Indeed , the results seen here indicate that dog ownership is the most clearly established risk factor for acquiring human AE . It seems possible that a domestic cycle of E . multilocularis exists at least in some regions , which may facilitate transmission of the parasite to humans and therefore increase the risk of human AE . The presence of E . multilocularis-infected dogs seems to play a central role for the occurrence of human AE in China [36] , while foxes , if present in affected regions , can act as additional definitive host for the parasite . In Europe , the role of dogs is less clear . Although it is undisputable that dogs may play a role in AE transmission in Europe , dog-related PRFs could represent confounders in areas where transmission might occur mainly through infected foxes . In these predominantly rural areas , people may be more frequently exposed to dog-related PRF , but the infection pressure might in fact be increased due to the presence of a large biomass of infectious E . multilocularis eggs excreted by foxes into the environment . However , we cannot exclude potential pathways of transmission related to the behavior of dogs , even when they are not infected . For instance , dogs commonly engage in a behavior known as scent-rolling: On finding any source of strong or unique-smelling , such as urine or faeces ( for instance dispersed by positive foxes ) or any other pungent odor that is not a regular scent within their territory , they roll their face and body in it , thus transferring the odor to their coat . Such behavior can increase the probability of eggs sticking onto their fur . We hypothesize that non-infected dogs roaming freely in endemic areas can act as carriers of E . multilocularis eggs , thus increasing the potential risk of human infection . It should also be noted that “being a dog owner” probably comprehensively includes variable human/dog behaviours reflecting a number of different PRFs . For instance , Kern and colleagues [19] identified an increase in odds of AE infection depending on factors such as “leaving the dog in the garden unattended” , “dog killing the game” , “allowed dog into the house“ , “walked dog without leash“ , “dog ate mice”and “had dog dewormed infrequently or never“ . The use of tailored questionnaires in endemic areas focusing on specific PRFs related to dogs in order to elucidate pathways of transmission mediated by human/dog behaviour , are strongly encouraged . Cats have occasionally been found infected with E . multilocularis , and a zoonotic potential cannot be totally ruled out , however they are considered as less suitable definitive hosts [37] . In a recently published study , a high worm burden was detected in cats , but it comprised only of immature worms , which indicated that the fecundity of E . multilocularis in cats was rather low [38] . In fact , E . multilocularis eggs derived from adult tapeworms harboured by cats failed to develop into metacestodes when used to experimentally infect mice [1] . More recently , faeces , identified to be of feline origin using genetic analysis , was found to contain eggs of E . multilocularis [39] . Additionally , cats may behave in a way that can increase the probability of transmitting the infection to humans , e . g . by E . multilocularis eggs sticking onto their fur . Interestingly , a significant increase in the risk of human AE have been identified for the cat-associated PRFs “left unattended outdoors” and “ate mice” [19] . Food-related risk factors such as eating mushrooms , wild berries or unwashed strawberries illustrate that vegetables and fruit , growing close to the ground may become contaminated with E . multilocularis eggs shed by infected definitive hosts , which may lead to human exposure to E . multilocularis if eaten raw or undercooked . The same may hold true for water , especially if contaminated surface water is used for drinking . The relative importance of food- or waterborne AE infection among other risk factors will depend , however , on general sanitation practices , e . g . washing wild berries or adequately treating natural surface water . The debate regarding the importance of unwashed contaminated fresh fruit , vegetables and mushrooms in the transmission of E . multilocularis is still ongoing [40–42] . Our meta-analysis illustrates that the chance of AE transmission through the ingestion of food and water contaminated with E . multilocularis eggs does indeed exist , but it is important to note that food- and water-borne PRFs do not significantly increase the risk of infection . Similar findings have been recently reported for CE [11] , although it should be emphasized that water- and food-borne transmission seems to be more evident for AE as compared to CE . This finding is also supported by a recent SR on foodborne parasitic diseases , in which the percentage of foodborne CE and AE ( reported as foodborne_DALYs/total_DALYs x100 ) was estimated to be 21% and 48% , respectively [43] . PRFs related to working or recreational activities in endemic areas may also point to an increased exposure of people to E . multilocularis eggs present on the fur of infected definitive hosts or in the environment , in particular in gardens and in areas used for agricultural purposes . In fact , not only “being a farmer” or “went to forests for vocational reasons” seems to be a significant PRF but also activities increasing the probability of ingesting eggs inadvertently such as “chewed grass” and “hunting / handling foxes” . The fact that particular HLA types appeared to be a potentially protective factor against AE infection may indicate that the susceptibility of humans to infection with E . multilocularis has a genetic component and that humans with these HLA antigens may be less likely to contract the disease or to develop clinically apparent AE [44] , but this aspect may have to be studied in more detail . Some socio-cultural factors ( for example , poverty as illustrated by low income and low level of education or ethnic background such as the increased risk of Tibetans of contracting AE ) may also play an important role in AE transmission . The literature research carried out in this study identified several publications , which contained information on PRFs presented in formats that precluded their extraction or the computation of OR . These PRFs included human immunosuppression [45] , landscape composition [46 , 47] , landscape and climatic characteristics [48] , age , occupation ( housewives ) , dog-ownership , consumption of wild vegetables and fruit [49] , farming [50–52] , presence of stray dogs and stray cats [51] , hunting [53] , changes in the ecology of wild hosts [54] and family clusters [55] . In addition , approaches used in meta-analysis may lead to the detection of statistically significant PRFs , which may nevertheless be spurious . This is why we explicitly refer to “potential” risk factors . Despite existing limitations of the SR approach , e . g . potential bias due to exclusion of studies that contain information in a format that is not compatible with SR or meta-analysis and the “pooling” of data from different studies , which may not always be sufficiently comparable , SR and meta-analysis are generally accepted as a tool in evidence-based science . In this literature review , age , gender and dog ownership have been identified as PRFs , however it should be emphasised that some of these may represent confounders . A critical discussion on confounding , intrinsic limits and advantages of case-control and cross-sectional studies on CE , that can be applied to AE , has been recently published [11] . It is also noteworthy that including data originating from different geographical regions ( for example from China and Europe ) in a meta-analysis , harbours risks and uncertainties , which must be taken into account when interpreting the results . Indeed , factors for acquiring AE in Europe might differ from those in central Asia , which may be due for example to different living conditions ( hygienic conditions in rural areas , income etc . ) and habits ( most European citizens are more or less sedentary and keep dogs or cats as pets , which may reduce the risk to prey on infected intermediate hosts ) . The majority of case-control ( 4/6 ) and cross-sectional ( 11/12 ) studies included in this SR originated from Europe and China , respectively . It was therefore reasonable to suspect that some of the PRFs that were significantly associated with an increase in the risk of developing AE may reflect certain sociocultural determinants . It should also be stressed that results obtained in the meta-analysis could be biased by the fact that a large proportion ( 91% ) of the global burden of AE originates from China [9] . In order to clarify this , we performed a meta-analysis , in which we compared case-control sub-studies from Europe ( n = 6 ) with those from all regions ( Europe , n = 6; Alaska , n = 1; China , n = 1 ) and cross-sectional sub-studies from China ( n = 12 ) with those reported globally ( China , n = 12; Europe , n = 1 ) . Since the results of these analyses suggested that relevant PRFs remained the same both in Europe and China , we decided to aggregate the data for these geographical areas , thus increasing the power of the statistical analysis . The absence of a difference between the PRFs observed here may , however , be related to the disproportion in the number of individuals enrolled in the studies originating from these areas . Our SR is limited to articles published in English , German , French , Polish , Finnish , Dutch , Spanish or Italian . Papers published exclusively in Chinese , Russian and Turkish ( i . e . without an abstract in any of the languages listed in the manuscript ) were not taken into account , since these languages were not understood in the consortium . However , relevant peer-reviewed studies are normally published in English or include an English abstract . Although Chinese was excluded from the language search , the highest number of retrieved studies used for data extraction and meta-analysis were from China ( n = 9 ) , but they were published in English . Furthermore , a study on the effect of language restriction on SR-based meta-analyses in conventional medicine found no evidence of bias as a result of language restriction [56] . This SR provides an up-to-date account of PRFs for human AE based on the meta-analyses of the available literature since 1950 . This study may constitute the basis for designing and improving programmes aimed at the control and prevention of this severe human disease .
|
Human alveolar echinococcosis is a severe zoonotic disease caused by the metacestode stage of the tapeworm Echinococcus multilocularis . The objective of this systematic literature review was to identify and summarize the current knowledge on potential risk factors associated with human alveolar echinococcosis . The categories of potential risk factors included dog-related factors such as dog ownership or play with dogs; vocational factors like being a farmer or handling foxes; human habits such as chewing grass; gender ( being female ) and socio-cultural factors like being Tibetan or having a low income or poor education , which may be relevant only in particular endemic areas . The identification of potential risk factors may help identify strategies that aim to decrease human infection with E . multilocularis .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"cross-sectional",
"studies",
"tropical",
"diseases",
"vertebrates",
"database",
"searching",
"parasitic",
"diseases",
"dogs",
"animals",
"mammals",
"research",
"design",
"mathematics",
"statistics",
"(mathematics)",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"echinococcosis",
"foxes",
"mathematical",
"and",
"statistical",
"techniques",
"pathogenesis",
"helminth",
"infections",
"cats",
"host-pathogen",
"interactions",
"meta-analysis",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"amniotes",
"statistical",
"methods",
"organisms"
] |
2017
|
Potential risk factors associated with human alveolar echinococcosis: Systematic review and meta-analysis
|
Effective regulation of primary carbon metabolism is critically important for bacteria to successfully adapt to different environments . We have identified an uncharacterised transcriptional regulator; RccR , that controls this process in response to carbon source availability . Disruption of rccR in the plant-associated microbe Pseudomonas fluorescens inhibits growth in defined media , and compromises its ability to colonise the wheat rhizosphere . Structurally , RccR is almost identical to the Entner-Doudoroff ( ED ) pathway regulator HexR , and both proteins are controlled by the same ED-intermediate; 2-keto-3-deoxy-6-phosphogluconate ( KDPG ) . Despite these similarities , HexR and RccR control entirely different aspects of primary metabolism , with RccR regulating pyruvate metabolism ( aceEF ) , the glyoxylate shunt ( aceA , glcB , pntAA ) and gluconeogenesis ( pckA , gap ) . RccR displays complex and unusual regulatory behaviour; switching repression between the pyruvate metabolism and glyoxylate shunt/gluconeogenesis loci depending on the available carbon source . This regulatory complexity is enabled by two distinct pseudo-palindromic binding sites , differing only in the length of their linker regions , with KDPG binding increasing affinity for the 28 bp aceA binding site but decreasing affinity for the 15 bp aceE site . Thus , RccR is able to simultaneously suppress and activate gene expression in response to carbon source availability . Together , the RccR and HexR regulators enable the rapid coordination of multiple aspects of primary carbon metabolism , in response to levels of a single key intermediate .
Soil-dwelling Pseudomonas spp . are exposed to a complex and dynamic physical and chemical environment , and must constantly adapt their cell physiology by changing the expression patterns of membrane proteins , secreted small molecules , and enzymes [1] . This ability to gauge the surroundings and modulate gene expression accordingly is crucial for effective environmental adaptation [2] . Therefore , the capacity of Pseudomonas to prosper in different niches depends not only on the acquisition of nutrients and effective resistance to external stresses , but also on the effective deployment of sensory proteins , signal-transduction pathways and transcriptional regulators [3] . While much research into niche colonisation by microbes has focussed on phenotypic adaptations such as motility , biofilm formation and stress resistance [4 , 5] , the remodelling of central metabolism to optimally respond to the external environment is a critical and understudied trait . The efficient coordination of carbon uptake and flux through the primary metabolic pathways with the nutrient availability of the surroundings is key to the successful colonisation of most ecological niches [6] , and relies on multilevel control of the expression and activity of proteins involved in primary metabolism [7] . A range of transcriptional factors orchestrates bacterial gene expression in response to carbon source availability , generating a sophisticated control network for both catabolic and anabolic metabolism [8] . Transcription factors belonging to the RpiR family control the expression of enzymes involved in carbon metabolism [9] . Members of this family are characterized by a helix-turn-helix DNA-binding domain at the N-terminus and a sugar isomerase domain ( SIS ) at the C-terminus [10 , 11] . In Pseudomonas , the RpiR regulator HexR controls the uptake and the catabolism of glucose , modulating the expression of glucose phosphorylative pathway and Entner-Doudoroff ( ED ) pathway genes [12–14] . HexR gene targets are grouped in two well-conserved operons containing genes for glucose catabolism and transport [14] . HexR binds as a dimer to the pseudo-palindromic DNA consensus sequence 5’- TTGTN7-8ACAA-3’ , found in the zwf/hexR and edd/gap-1 intergenic regions , and represses gene transcription in its apo-form . When glucose is present , it is imported and metabolised via several steps to 2-keto-3-deoxy-6-phosphogluconate ( KDPG ) , an exclusive metabolic intermediate of the ED pathway . HexR binds KDPG via its SIS domain , signalling glucose availability and causing HexR to dissociate from DNA , inducing expression of its gene targets and stimulating glucose uptake and metabolism [14] . Downstream of glycolysis , the Krebs cycle is responsible for the complete oxidation of acetyl-CoA to CO2 and provides intermediates that are necessary for the production of amino acids and other cellular macromolecules . The pools of these metabolic intermediates must be constantly resupplied to maintain them at sufficient levels for metabolism and growth . For this reason it is important that flux through and around the Krebs cycle is exquisitely controlled to make best use of available carbon sources [15] . During Krebs cycle operation , acetyl-CoA condenses with oxaloacetate to form citrate , which is , following a complete turn of the cycle , reconverted to oxaloacetate . Each turn of this cycle involves the loss of two molecules of CO2 . When acetyl-CoA is the only available carbon source , classical Krebs cycle operation cannot assimilate carbon . Consequently , when acetate or fatty acids are the primary source of carbon and energy , many bacterial species including Pseudomonas [16 , 17] activate a specific anaplerotic pathway , called the Glyoxylate shunt [18] . The Glyoxylate shunt is common among microorganisms and higher plants [19 , 20] , and diverts part of the carbon flux at isocitrate [18 , 21] . The key enzymes of this pathway are isocitrate lyase ( ICL ) , which converts isocitrate to glyoxylate , and malate dehydrogenase , which condenses a molecule of glyoxylate with an acetyl-CoA to produce malate and succinate . This ensures the replenishment of the metabolic intermediates required for the biosynthesis of cellular components . The only regulatory mechanism for the Glyoxylate shunt known to date involves the reversible phosphorylation of isocitrate dehydrogenase ( IDH ) [22–24] . When bacteria grow on two-carbon compounds , such as acetate , IDH phosphorylation inactivates this enzyme and forces carbon flux through the shunt pathway . P . fluorescens is a common soil bacteria that non-specifically colonizes plant roots , and can improve plant health through nutrient recycling , pathogen antagonism , and inducing plant defence responses [3] . The rhizosphere environment is both highly complex and attractive to microbial life . Plant roots continuously produce and secrete compounds in the rhizosphere environment ( including ions , amino acids , organic acids , sugars , phenolics and other secondary metabolites ) [25 , 26] , and many microorganisms are attracted to these exuded nutrients [27] . Successful root colonisation depends on both the coordinated expression of factors involved in phenotypes such as motility and biofilm formation [5 , 28] and on the adaptive remodelling of central metabolism [4 , 29] . Recently , an uncharacterised member of the RpiR family; rccR ( PFLU6073 ) , was identified in an In Vivo Expression Technology ( IVET ) screening experiment for loci involved in P . fluorescens plant interactions [30] . RccR shares a very high amino-acid sequence identity ( 43% identical ) with HexR , and both transcription factors play important roles in wheat rhizosphere colonisation and growth on defined carbon sources . Furthermore , the HexR/RccR regulon is widespread among the pseudomonads and in several other bacterial genera . This prompted us to investigate the contributions of these ‘twin’ transcription factors to environmental adaption in more detail . We first examined the role of HexR in P . fluorescens during the wheat rhizosphere colonisation , confirming its function as a regulator of glucose metabolism , consistent with earlier data [14] . Next , we characterized RccR , and identified it as a master regulator of pyruvate metabolism , the glyoxylate shunt and gluconeogenesis . RccR binds to two distinct , pseudo-palindromic binding sites in the promoter regions of its target genes , which share a binding sequence but differ in the length of the linker region . RccR shares a ligand; the ED pathway intermediate 2-dehydro-3-deoxy-phosphogluconate ( KDPG ) , with HexR , reinforcing the regulatory connection between the two proteins . Furthermore , ligand binding increases RccR affinity for one binding site but decreases it for the other , enabling a sophisticated transcriptional response to a single metabolic intermediate . We propose that the coordinated activity of HexR and RccR tightly controls the remodelling of central metabolism in response to intracellular KDPG levels , maximizing bacterial proliferation and fitness , and optimising enzyme synthesis to most effectively respond to nutrient availability in the environment .
The P . fluorescens SBW25 rccR ( PFLU6073 ) gene encodes an uncharacterised member of the RpiR family of transcriptional regulators . Following the observation that rccR seems to play a role in the plant environment [30] we decided to examine the locus in more detail . SBW25 RccR has a striking degree of similarity ( 43% amino acid identity , over 70% similarity ) to the central metabolic regulator HexR . Examination of the primary structure of both proteins indicated that the key residues of the DNA binding motif ( R57 , R60 ) , and ligand binding site ( S139 , S183 ) are conserved in both cases ( Fig 1A ) . To identify any differences between the two proteins , molecular models of both were produced based on four published crystal structures of transcriptional regulators ( PDB files: 3sho , 2o3f , 4ivn and 3iwf ) . Both models were essentially identical , with no obvious differences in binding sites or residue conformation emerging from this analysis ( Fig 1B ) . Despite their highly similar predicted structures , hexR does not appear in the initial sugar beet IVET dataset [30] , suggesting distinct regulatory roles for RccR and HexR . To test this , we examined the significance of both rccR and hexR to rhizosphere colonisation and plant interaction . Single and double rccR/hexR mutants were produced in P . fluorescens SBW25 , and a series of competitive colonisation assays performed with lacZ-labelled WT SBW25 . After seven days , significantly fewer ΔrccR , ΔhexR and ΔrccRΔhexR colony forming units ( CFUs ) were recovered from model rhizospheres compared with the WT-lacZ competitor ( Fig 1C ) , indicating that both genes are similarly important for competitive growth in the plant root environment . We examined the responses of the rccR and hexR mutants to different nutrient conditions by testing their ability to grow in rich , complex and defined minimal media containing different sources of carbon . We observed little effect of rccR/hexR gene deletion in rich media; no growth-rate difference in KB , and a slight delay in log phase entry for both mutants in LB ( Fig 2A/2B ) . Conversely , deletion of rccR resulted in severely compromised bacterial growth in minimal medium with either glucose or glycerol as the sole carbon source ( Fig 2C/2D ) . Deletion of hexR on the other hand affected growth in media containing pyruvate , acetate or succinate , ( Fig 2E , 2F and 2G ) . These growth defects were dominant and not additive: the ΔrccR/ΔhexR double mutant displayed an rccR-mutant phenotype in glycerol and glucose , and a hexR-mutant phenotype in pyruvate , acetate , or succinate ( Fig 2 ) . Interestingly , an exception is the bacteria growth in acetate , in which the double mutant shows an intermediate ability to grow compared to the wild type and the single mutants ( Fig 2F ) . Despite their marked sequence and structural similarities , the activities of RccR/HexR are clearly distinct , with their relative importance varying markedly depending on the available carbon sources . Previous studies have shown that HexR is a transcriptional repressor of the glucose phosphorylative and Entner-Doudoroff pathways in P . putida and P . aeruginosa [14 , 31] . HexR represses the expression of its gene targets by binding to the specific consensus sequence 5’- TTGT-N7-8-ACAA-3’ ( Fig 3A ) . To verify the HexR regulatory system ( Fig 3B ) in P . fluorescens , we first examined the intergenic regions between the p-edd and p-gap-1 , and between the p-zwf and p-hexR promoters in the SBW25 genome for the HexR consensus binding sequence . As expected , both intergenic regions contained HexR binding sites . Next , qRT-PCR experiments were performed to examine the expression of the gap-1 , edd and zwf genes in WT and ΔhexR backgrounds , for bacteria grown in minimal media containing different , single carbon sources . While little difference in expression of these genes was observed between WT and ΔhexR for bacteria grown in glucose minimal media ( Fig 3C ) , the expression profile of the ΔhexR mutant significantly diverged from WT when the strains were grown in glycerol , pyruvate or acetate media . Under these conditions , expression of gap-1 , edd and zwf significantly increased in the ΔhexR background ( Fig 3D , 3E and 3F ) . These results suggest the strong repression of gap-1 , edd and zwf genes by P . fluorescens HexR , particularly under conditions where KDPG is unable to accumulate , and consistent with earlier data [14] . To test whether HexR represses its own expression , we constructed a p-hexR-lacZ transcriptional reporter plasmid ( pME-hexR ) and performed β-galactosidase assays to study the promoter activity of hexR in WT SBW25 and the ΔhexR mutant , grown in either minimal glucose or pyruvate medium . We observed no autoregulation of hexR promoter activity when bacteria were grown in any of the tested media ( Fig 3G ) . However , qRT-PCR data showed a higher hexR expression level when grown with glucose or glycerol , but not with pyruvate or acetate as the sole carbon source ( Fig 3H ) . Despite the high predicted structural similarity between HexR and RccR , their markedly different impacts on SBW25 growth suggest that these proteins control independent transcriptional regulons . To characterise the RccR regulon , we performed a ChIP-seq assay using a polyclonal RccR antiserum on SBW25 WT/ΔrccR strains grown in minimal pyruvate and glycerol media . While little difference was seen between the two media conditions , based on a stringent peak-calling analysis we were able to identify 8 RccR binding sites from this experiment ( Fig 4 ) , including one in the rccR promoter region itself ( Fig 4G ) . The peaks identified in our RccR ChIP-seq assays corresponded to strongly enriched regions relative to the respective ΔrccR controls . All 8 peaks were localized in intergenic regions , upstream of one or more genes ( Table 1 ) . To verify that the identified binding sites are associated with RccR-mediated regulation , we extracted RNA from SBW25 WT and ΔrccR grown in M9 glycerol , and performed qRT-PCR assays for each member of the putative RccR regulon ( except rccR ) . Where a putative binding site lay between two divergent genes , both targets were tested and RccR regulation shown unambiguously for one of them ( S1 Fig ) . In each case , we saw a marked increase of expression in ΔrccR relative to WT for only one of the two possible targets ( S1 Fig ) , indicating RccR repression . The RccR regulon includes enzymes involved in gluconeogenesis ( phosphoenolpyruvate carboxykinase ( pckA ) and glyceraldehyde-3-phosphate dehydrogenase ( gap ) , NADPH/NAD+ redox balance ( NAD ( P ) transhydrogenase ( pntAA/PFLU0112/pntB ) , acetyl-CoA production ( aceE/F ) and the glyoxylate shunt pathway ( isocitrate lyase ( aceA ) and malate synthase G ( glcB ) ) ( Table 1 ) . Most members of the RccR regulon have well-understood enzymatic functions in carbon metabolism . The notable exception , PFLU2154 , encodes a hypothetical protein of unknown function . This gene shows a similar expression profile to aceA , with particularly strong RccR repression seen in glycerol , and much less effect in acetate ( Fig 5A/5D ) . This suggests that this gene may play an important role in two-carbon substrate catabolism . To test this , a non-polar PFLU2154 deletion mutant was produced and its ability to grow on different carbon sources was tested . Consistent with both our hypothesis and previous data , the only condition in which the mutant showed a growth defect was in acetate medium ( S2A/S2H Fig ) . We further examined the ability of the ΔPFLU2154 mutant to competitively colonize wheat roots . No significant differences in colonisation were seen for the mutant compared to WT SBW25 ( S2I Fig ) . To more closely examine the role of RccR in SBW25 gene regulation , we extracted additional RNA samples from WT and ΔrccR cultures grown in minimal media containing glucose , pyruvate or acetate , and used qRT-PCR to probe expression of the target genes in Table 1 . The impact of rccR deletion on gene expression broadly matched its impact on growth in different carbon sources . When grown in glucose or glycerol as the sole carbon source , RccR tightly repressed the aceA and glcB genes , indicating inhibition of the glyoxylate shunt pathway . On the other hand , much less control was seen for aceE/F gene expression ( Fig 5A/5B ) . When bacteria grew in minimal pyruvate the aceA and glcB genes are still repressed but to a lesser ( albeit still significant ) extent than in glucose and glycerol . Again , little repression was seen for aceE/F ( Fig 5C ) . A much more striking shift was seen for strains grown in minimal acetate media ( Fig 5D ) . In this case , no inhibition was observed for either aceA or glcB gene expression , suggesting activation of the glyoxylate bypass in these conditions . While rccR deletion in acetate media had little effect on most tested genes , the aceE/F locus was strongly expressed ( Fig 5D ) , indicating a strong down-regulation of pyruvate dehydrogenase production under conditions where pyruvate metabolism is no longer required and acetyl-CoA may be produced directly from acetate . To probe RccR control of its own transcription , we cloned the rccR promoter region into the pMElacZ plasmid , and examined the impact of rccR deletion on β-gal activity in different carbon conditions . As expected , RccR repressed its own expression , with increased β-gal activity in the ΔrccR strain in every condition tested: glucose , glycerol , pyruvate and acetate minimal media ( Fig 5E ) . Consistent with this , qRT-PCR data showed an approximately constant rccR gene expression when grown in glucose , glycerol or pyruvate medium , although a lower expression was observed with acetate as the sole carbon source ( relative to carbon-free rooting solution ) ( Fig 5F ) . The sequence surrounding the eight RccR ChIP-seq binding sites was analysed for potential binding motifs by MEME . A 28 bp pseudo-palindromic consensus sequence was found in 7 sites , with an E-value of <1 . 8e-10 and p-value <0 . 05 for each sequence ( Fig 6A ) . The 28 bp pseudo-palindrome was not identified in the upstream regions of pckA or aceE . The ChIP-seq dataset for pyruvate grown cells contains an additional , smaller peak inside the pckA open reading frame ( Fig 4B ) , corresponding in location to a second 28 bp pseudo-palindrome located around 300 bp after the start codon ( Fig 6A ) and suggesting an additional layer of RccR-regulation for this gene . Subsequent manual analysis of the pckA promoter located a 29 bp RccR-pseudo-palindrome with an additional base-pair in the linker region ( Fig 6A ) , likely accounting for the failure of MEME to identify it . In the case of aceE , the upstream region contains two copies of a single 15 bp pseudopalindromic sequence , separated by 68 bp . These two sequences ( 15 bp and 28 bp ) contain the same palindromic repeat ( TGTAGT/ACTACA ) , with the only difference between them the length of the inter-repeat region ( Fig 6B ) . The ChIP-seq data for the aceE upstream region contains a distinctive double peak in both tested conditions ( Fig 4C ) , suggesting that RccR binding occurs to both 15 bp sites in the aceE upstream region . Interestingly , in the rccR upstream region we identified a degenerate 28 bp consensus sequence , wholly containing the shorter 15 bp sequence ( Fig 6B ) . To verify the RccR binding on these predicted consensus sequences , we performed ReDCaT SPR assays [32] using C-terminal His-tagged purified protein ( RccR-His ) . We confirmed the interaction between RccR-His and the predicted RccR binding sites for every tested target , with weaker RccR binding seen for the pckA , pntAA and gap sequences compared to the other RccR targets ( Fig 7A ) . We saw only very weak RccR binding to pckA* , which combined with its intragenic location suggests that this sequence may not represent a relevant RccR binding target . Next , in order to more closely analyse RccR binding to the two distinct consensus sequences , we performed further SPR experiments to study RccR-His binding affinity to aceE , aceA , and rccR . These analyses showed that RccR binds to the rccR binding site , and to a single copy of the 15 bp aceE sequence ( aceE > rccR ) , with considerably higher affinity than to the aceA sequence ( Fig 7B ) . In order to verify and map the predicted RccR binding sites , we first performed 5’ RACE analysis of the strongly binding DNA targets to identify transcriptional start sites ( S3 Fig ) . Next , we performed DNaseI footprinting experiments on the rccR , aceA and aceE promoter regions using the RccR-His protein at an nM concentration range . Fig 8A indicates that RccR specifically binds to all the tested targets . A single binding region was observed for rccR and aceA ( Fig 8A , lanes 1 to 6 and lanes 7 to 12 , respectively ) , with the latter binding site containing a central DNaseI hypersensitive band . Notably , for the aceE fragment , we found two distinct binding sites flanking a DNA region rich in hypersensitive bands , suggestive of DNA bending ( Fig 8A , lanes 13 to 18 ) . Under our experimental conditions , RccR seems to recognize both aceE and rccR with very high relative binding affinity ( aceE > rccR , with full protection observed at 10 nM and 20 nM RccR-His on aceE and on rccR , respectively ) , while aceA is bound with a lower affinity ( full protection at 80 nM RccR-His ) . Given the apparent binding cooperativity between the two aceE 15 bp sites , it is possible that the binding affinity of RccR for the aceE upstream region in vivo is even higher than that calculated by SPR for a single aceE site ( Fig 7B ) . The RccR binding sites for all three sequences were precisely mapped ( reported in Fig 8B ) and shown to fully overlap the MEME-predicted palindromic binding sequence . Moreover , RccR binding to the genomic fragments tested occurs in the core promoter regions , encompassing the verified transcriptional start sites on p-rccR and p-aceE . To further characterize RccR-DNA interactions , we carried out hydroxyl-radical footprinting assays ( OH-FP ) . In this technique , DNaseI is substituted by radical ions as cutting agents to obtain a higher resolution ( S4A Fig ) . For all three tested promoter probes , RccR binding resulted in a pattern of short protected tracts of 3/4 nucleotides in length , separated by non-protected regions . As before , nucleotides protected in the OH-FP experiment were mapped onto the promoter sequence ( S4B Fig ) . Intriguingly , for all the tested binding sites , the nucleotides protected in OH-FP belong to the spacer that separates the conserved repeats of the consensus binding motif , or falls immediately upstream/downstream of the conserved repeats ( S4B Fig ) . Regions protected in hydroxyl-radical footprinting experiments reflect the minor accessibility of radical ions to the DNA minor groove , and for this reason these protected regions do not necessarily represent the portions of the probe directly contacted by the protein . Considering that RccR footprint regions surround the inverted repeats , these data are consistent with RccR interaction with the conserved repeats in the DNA major groove , narrowing the adjacent minor grooves and protecting these sequences in vitro from hydroxyradical digestion . Additional binding experiments are required to fully characterise the RccR-DNA complex . Nonetheless , these data reinforce the crucial role of the conserved sequences in mediating specific RccR binding . The RccR binding pattern on aceE is particularly complex ( S4A Fig , lanes 9 to 12 ) : there are 4 short protected stretches close to the conserved sequences ( black boxes in S4A Fig and black dots in S4B Fig ) , and 2 additional protected tracts in the DNA region that separates the two binding sites ( grey boxes in S4B Fig and grey dots in S4B Fig ) . Because this is a probe harbouring two separate operators bound to RccR , this suggests that the DNA fragment may undergo strong bending or a similar perturbation to enable RccR binding . The existence of two different RccR consensus binding sequences could explain why the pattern of regulation for aceE and rccR in different carbon sources differs markedly from the rest of the RccR regulon . As a member of the RpiR protein family , RccR contain an SIS domain for binding an effector-sugar . The high sequence homology between HexR and RccR and the conserved amino acids residues in the SIS domains , suggest that the RccR hypothetical effector could be a phosphorylated sugar , more likely an intermediate of the central carbon metabolic pathway controlled by both HexR and RccR . In order to identify the RccR ligand we performed SPR experiments , using the ReDCaT system , that allowed us to analyse the RccR-His binding on the aceE , aceA and rccR consensus sequences immobilised on the ReDCaT Chip , in presence of specific metabolic intermediates . No response was seen for any of the tested sugars ( S5 Fig , Fig 9 ) , with the exception of the well-characterised HexR ligand KDPG ( Fig 9 ) . KDPG addition decreased the affinity of RccR-His for both aceE and rccR consensus sequences in the μM concentration range . Conversely , RccR binding affinity for the aceA ( 28bp ) binding site increased in-line with a rising KDPG concentration ( Fig 9 ) . As such , it appears that KDPG is the specific RccR effector , and is responsible for the induction of the aceE/F operon and the conversion of pyruvate into acetyl-CoA . In the presence of KDPG , rccR auto-repression is reduced by a modest amount , in line with a shift from the 15 bp binding site to the 28 bp consensus on KDPG binding . Partial release of rccR repression , alongside increased binding affinity for the 28 bp sites upstream of the target genes , both serve to repress glyoxylate shunt/gluconeogenesis activity when KDPG levels are high . To assess the importance of RccR regulation in other bacterial species , a BLAST analysis with the SBW25 rccR gene sequence was conducted on 1502 publically-available bacterial genomes . Because of the high sequence identity shared by rccR and hexR , we performed a reciprocal BLASTp screen to discard false positive hits ( representing hexR or other rpiR family members ) . This second step allowed us to confidently identify rccR homologs in numerous pathogenic and non-pathogenic Pseudomonas spp . and in dozens of additional bacterial genera ( S1 Table ) . We next extended our analysis to examine the conservation of the RccR regulon throughout the Pseudomonas genus . A search of all publically-available Pseudomonas genomes was conducted using the consensus 28 bp and 15 bp pseudo-palindromic binding sequences , with stringent parameters to avoid false-positives ( S2 Table ) . Multiple occurrences of the RccR consensus binding sequences were identified in almost every species in the genus . In every case , these binding sites were located upstream of genes found in the SBW25 RccR regulon , supporting a widespread and highly conserved role for RccR in the control of bacterial carbon metabolism and glyoxylate shunt regulation . Given the widespread conservation of rccR and its importance for carbon utilisation and growth in P . fluorescens , we examined the similarity of the SBW25 regulatory network with that of the opportunistic human pathogen P . aeruginosa PAO1 . The P . aeruginosa PA01 genome encodes a close homolog of rccR alongside most of the confirmed RccR gene targets identified in SBW25 ( S3 Table ) . The only significant variation between the two regulons is the absence of a PFLU2154 homolog in PA01 . Manual examination of the upstream regions of the conserved genes in PA01 confirmed the presence of predicted RccR binding sequences in every case . Moreover , qRT-PCR expression analysis of the rccR gene targets in PA01 , for cells grown in glycerol , pyruvate and acetate as the sole carbon source confirmed that RccR controls glyoxylate shunt and gluconeogenesis gene expression in P . aeruginosa in a similar fashion to P . fluorescens ( S6 Fig ) .
In this study we identify and characterise a novel transcriptional regulator in Pseudomonas that regulates the expression of primary metabolic pathway genes in response to carbon source availability in the surrounding environment . The rccR gene was identified in a screen for up-regulated loci during P . fluorescens plant interaction [3] , and encodes an RpiR family protein with a remarkably high amino acid sequence identity to the glycolysis regulator HexR . Members of the RpiR protein family contain an N-terminal helix-turn-helix DNA binding domain and a C-terminal ‘SIS’ sugar isomerase domain . The predicted structures of RccR and HexR are highly similar , with the important residues for HexR DNA and ligand binding [14] conserved in RccR . This high degree of structural similarity initially suggested that RccR and HexR may share a common role in the regulation of carbon metabolism , possibly as competitors for a shared binding site . Thus , we decided to examine HexR alongside RccR in P . fluorescens SBW25 , and to determine the roles of both regulators during bacteria-plant interactions . Both ΔrccR and ΔhexR SBW25 mutants displayed significantly compromised ability to colonize the rhizosphere of wheat seedlings , suggesting that both regulators contribute to efficient bacterial growth in the root environment , presumably via transcriptional control of carbon metabolism in response to the available plant root exudates . However , while both regulators are similarly important for efficient root colonisation , the rccR/hexR mutant strains show markedly different growth effects in minimal defined media . HexR is required for efficient SBW25 growth on pyruvate or acetate or succinate as the sole carbon source . P . putida HexR functions as a transcriptional repressor of genes involved in the glucose phosphorylative and Entner-Doudoroff pathways [14] . When bacteria grow on two/three-carbon sugars ( or their precursors ) , HexR represses genes required for glucose transport and the initial stages of the Entner-Doudoroff pathway . Conversely , when glucose is available as a carbon source , these HexR gene targets are expressed . In the presence of glucose , the metabolic intermediate KDPG is produced and binds to the SIS domain of HexR . This leads to a release of DNA binding and the consequent transcriptional activation of the HexR regulon [14] . Based on our expression analysis of the ΔhexR mutant grown in different media conditions , P . fluorescens HexR clearly plays a similar role in regulating glucose metabolism as in other Pseudomonas spp . [14 , 31] Conversely , compromised growth of the ΔrccR mutant on glucose or glycerol suggests that RccR may work to produce an efficient metabolic response to an entirely different set of carbon sources . To understand the regulatory role of RccR in SBW25 we performed a ChIP-seq experiment that identified eight RccR binding sites , and a corresponding set of RccR gene targets . Firstly , RccR modulates its own expression , suppressing rccR gene expression in every condition tested , with the strongest repression seen for cells grown on acetate . Four of the RccR regulon genes; aceA , glcB , pntAA , and pckA are typically upregulated when bacteria grow in the presence of carbon sources that directly enter primary metabolism via acetyl-CoA , i . e . acetate , acetoacetate and fatty acids [33] . In agreement with this , our qRT-PCR analysis showed that these four genes are strongly repressed by RccR in glycerol and to a lesser extent in pyruvate , but not when acetate is the sole available carbon source . Isocitrate lyase ( aceA ) and malate synthase G ( glcB ) comprise the Glyoxylate shunt pathway , whose activation during growth on two-carbon molecules enables the replenishment of metabolic intermediates during Krebs cycle operation ( Fig 10 ) . The NADH/NAD+ cofactor pair plays a major role in microbial catabolism and it is crucially important for continued cell growth that NAD ( P ) H be oxidized to NAD ( P ) + and a redox balance be achieved [34 , 35] . The pntAA/0112/B operon encodes a 3-subunit transhydrogenase enzyme that catalyses the reaction NADPH + NAD+ ⇄ NADP+ + NADH , based on the intracellular NADPH/NAD+ ratio . Finally , phosphoenolpyruvate carboxykinase ( pckA ) is involved in the first step of the gluconeogenesis pathway , a necessary anabolic pathway for bacterial growth on acetate and other two-carbon sources [36] ( Fig 10 ) . Of the remaining RccR gene targets , two are suppressed during growth on glucose , glycerol or pyruvate , but less so on acetate . The gap gene encodes for glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) . Interestingly , RccR and HexR appear to regulate the expression of two different isoenzymes , gap ( PFLU1566 ) and gap-1 ( PFLU4965 ) working at the same metabolic step . However , these two isoenzymes have different roles in gluconeogenesis and hexose catabolism respectively , according to the relative levels of their preferred substrates ( NAD+ and NADP+ ) [37] . PFLU2154 encodes a hypothetical protein of unknown structure and function , and is particularly strongly repressed in glycerol medium . This gene is apparently not essential , as it can be deleted in SBW25 and is not present in every Pseudomonas species . The defective growth of the ΔPFLU2154 mutant strain in acetate medium suggests a role for this protein in either two-carbon source metabolism or gluconeogenesis , consistent with our expression data for the RccR regulon . The final RccR gene target is the aceE/F operon , which presents an entirely different regulatory pattern to the other members of the RccR regulon . These genes , encoding two subunits of a pyruvate dehydrogenase , are expressed in the presence of glucose , glycerol and pyruvate but are strongly repressed in acetate . This makes sense in the context of a switch from glycolysis to two-carbon metabolism and gluconeogenesis . AceEF suppression effectively prevents the further metabolism of pyruvate to acetyl-CoA , enabling the accumulation of carbon above this stage of glycolysis by PckA ( Fig 10 ) . Clearly , when grown on acetate as the sole carbon source , RccR enables both the repression and de-repression of different gene targets to produce an integrated regulatory response . MEME analysis of the sequences surrounding each of the 8 RccR binding sites presents a first step towards understanding this intricate regulatory system . A pseudo-palindromic , 28 bp consensus sequence was identified in the upstream regions of the seven glycerol/pyruvate suppressed RccR targets , while two copies of a shorter , 15 bp pseudo palindrome were found upstream of the aceE/F operon . The consensus sequence of these sites is the same ( TGTAGT/ACTACA ) , with the only difference the length of the internal linker . Subsequent SPR and Footprinting analyses confirmed that RccR binds to both of these consensus sequences , albeit with markedly different binding affinities . The presence of two distinct but related binding sites upstream of the RccR targets suggests that RccR exerts discrete regulatory effects on different genes in its operon through distinct DNA binding mechanisms . The 28 bp consensus sequence found upstream of rccR is particularly interesting in this context , as it contains the 15 bp consensus sequence within it . The presence of this double binding site upstream of the rccR gene could explain the constant repression observed for rccR in both acetate and glycerol/pyruvate media . Overlaying the RccR binding sites onto the six 5’ RACE-mapped promoter regions ( S3 Fig ) adds an additional layer of complexity to the RccR regulon . RccR binding overlaps ( or is very close to ) the +1 sites of pckA , aceE , glcB and rccR , implying a role for RccR in inhibition of promoter escape or inhibition of open complex formation on these promoters [38] . On the other hand , the RccR binding sites on the aceA and PFLU2154 promoters are located 40/41 bases upstream of the +1 site , suggesting a distinct repression mechanism applies for aceA and PFLU2154 . This would also be consistent with the much stronger conditional repression ( relative to pckA and glcB ) seen for these genes in our q-PCR experiments . Informed by the structural similarity between HexR and RccR , we confirmed the ED pathway intermediate KDPG as the sugar ligand for RccR . Identifying this RccR effector was the breakthrough that allowed us to interpret how the intricate HexR/RccR regulon controls primary carbon metabolism in Pseudomonas . In the absence of ligand binding , RccR-His has a far higher affinity for DNA sequences containing the 15bp consensus binding site than for the 28bp site . As repression of the aceEF operon only occurs in acetate-containing media , the affinity for this consensus sequence must decrease in conditions where pyruvate needs to be converted into acetyl-CoA . As predicted , KDPG binding to RccR markedly decreases binding affinity to the 15bp consensus , consistent with release of aceEF repression . Similarly , rccR expression is itself slightly de-repressed in the presence of the ligand . Conversely , KDPG binding RccR enhances binding to the lower-affinity 28bp binding sites upstream of aceA , pckA and other glyoxylate shunt/gluconeogenesis genes . Combined with the mildly increased expression of the RccR repressor itself , this results in strong inhibition of gene transcription from these loci when KDPG is present . This regulatory mechanism may help to explain the intermediate growth defect seen for the hexR/rccR double mutant grown on acetate media ( Fig 2F ) . In the other tested carbon conditions , the double mutant phenocopies the defective single mutant—i . e . the loss of transcriptional regulation of either the glyoxylate shunt or the ED pathway leads to inefficient metabolism and a growth defect . Most of the time , the additional loss of the non-repressing regulator is phenotypically neutral . In acetate , we propose that uncontrolled KDPG production in the hexR knockout leads to RccR inhibition and the loss of glyoxylate shunt function . Recovery of glyoxylate shunt activity in the double mutant allows the cells to use the available 2C carbon source , leading to a partial recovery of growth rate . By combining the activities of RccR and HexR , P . fluorescens controls multiple stages of primary carbon metabolism , from the glucose phosphorylative and Entner-Doudoroff pathways , to pyruvate catabolism , glyoxylate shunt and gluconeogenesis , by monitoring the concentration of a single metabolic intermediate , KDPG . Why the 15 and 28bp binding sites show such different affinities , and opposing responses to RccR ligand binding is currently unclear , but may be related to the length of the linker regions and the dimerization state of the transcriptional regulator . Because the length of a full DNA turn is 10 . 5 nucleotides , two RccR monomers binding to the shorter consensus can easily interact with each other on the same DNA face . Conversely , such interactions are likely to be more difficult on the 28bp consensus sequence , where our data suggest that DNA binding is required to enable RccR homodimer interaction . If KDPG binding changes the conformation of the RccR dimer , this may simultaneously stabilise the more flexible long-range complex and inhibit binding to the higher-affinity but less flexible 15bp site . A structural analysis of the RccR-DNA bound complex would be helpful for clarify the true mechanism of RccR function . Our analysis shows that the RccR regulator is conserved across multiple bacterial genera , and in every tested Pseudomonas species , supporting the importance of this transcription factor in the control of carbon metabolism . Given the strong similarity between RccR and HexR , it is reasonable to assume that these proteins are the product of gene duplication at some point in the past . Further evidence in support of this comes from the Shewanella oneidensis HexR homolog , which is proposed to control the ED and glucose phosphorylative pathways as well as the glyoxylate shunt and gluconeogenesis [39] . In Pseudomonas this role has clearly been subdivided between the “twin” proteins HexR and RccR . The RccR/HexR pathway enables P . fluorescens to dynamically fine-tune its metabolic pathways to quickly respond to carbon source availability . By using two transcriptional regulators with three different DNA target sequences to sense the concentration of a single key metabolic intermediate , RccR/HexR represents an elegant mechanism to enable environmental adaptation .
P . fluorescens strains were grown at 28°C , while P . aeruginosa and E . coli were grown at 37°C in lysogenic broth ( LB ) , solidified with 1 . 5% agar where appropriate . For growth experiments , bacteria were grown in M9 minimal media with single carbon sources added to a final concentration of 0 . 4% w/v , unless otherwise stated . Kanamycin ( Kc ) was used at 50 μg/ml , Gentamicin ( Gm ) was used at 12 . 5 μg/ml , Tetracycline ( Tet ) was used at 10 μg/ml for E . coli , 12 . 5 for P . fluorescens and 100 μg/ml for P . aeruginosa , while Carbenicillin ( Cb ) was used at 100 μg/ml . X-Gal was used at 40 μg/ml and IPTG at 100 μg/ml . All bacteria strains and plasmids used in this study are listed in S4 Table . Common molecular biology methods including plasmid DNA extraction , transformation , cloning , restriction digests , electrophoresis , purification of DNA fragments and sequencing were carried out according to standard molecular biology techniques as described previously [40] . PCR reactions were performed using GoTaq or Phusion DNA polymerase as appropriate . All oligonucleotides used in this study are listed in S5 Table and have been designed mainly on the basis of the P . fluorescens SBW25 ( GenBank NC_012660 . 1 ) and P . aeruginosa PAO1 ( GenBank NC_002516 ) genome sequences . P . fluorescens and P . aeruginosa deletion mutants were constructed via an adaptation of the protocol described elsewhere [41] . Up- and downstream flanking regions to the target genes were amplified by PCR using primers: 1–2 , 3–4 and 11–12 , 13–14 for rccR deletion in SBW25 and PAO1 respectively; 5–6 , 7–8 and for hexR deletion in SBW25; 17–18 , 19–20 for PFLU2154 deletion . The products in each case were ligated into pTS-1 between XhoI-BamHI . The resulting vectors were transformed into the target strain , and single crossovers were selected on Tet and re-streaked for single colonies . Cultures from single crossovers were grown overnight in 50 ml LB medium , then a dilution series of the overnight culture was plated onto LB plates containing 10% sucrose to enable counter-selection . Individual P . fluorescens colonies were patched onto LB plates ± Tet , with Tet-sensitive colonies tested for gene deletion by colony PCR using external primers . HexR and RccR sequence comparisons were carried out using the NCBI BLAST online tools ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) . The model protein structure of SBW25 RccR was constructed using HHPred ( Homology detection & structure prediction ) , followed by HMM-HMM comparison and MODELLER . The model generated was based on 4 published crystal structures ( PDB file names: 3sho , 2o3f , 4ivn , and 3iwf ) . To search for rccR by reciprocal BLAST using BLASTp , we randomly selected one species from each of the annotated bacterial genera available in the EBI collection of bacterial genomes ( https://www . ebi . ac . uk/genomes/bacteria . details . txt ) . This resulted in 1502 bacterial genomes being searched , with the results of the search shown in S1 Table . The sample pool contained one randomly selected representative of each publically available , annotated bacterial species , from a total available collection of 3794 genomes at NCBI . RccR transcriptional binding sites were identified by analysing the sequences surrounding the RccR ChIP-seq peaks using MEME ( Multiple EM for Motif Elicitation ) ( http://meme . nbcr . net ) [42] , with additional manual analysis of each region conducted as discussed in the results . All Pseudomonas genomes in the EBI collection of bacterial genomes were searched with the RccR binding site consensus by doing a simple string search ( using Perl ) allowing for 1 , 2 or 3 mismatches over the entire length of the consensus sequence . The results are shown in S2 Table . Paragon wheat seeds were sterilized with 70% ethanol and 5% hypochlorite , washed and germinated on sterile 0 . 8% MS agar for 72 h in the dark . Seedlings were then transferred into sterile 50 ml tubes containing medium grain vermiculite and rooting solution ( 1 mM CaCl2 . 2H2O , 100 μM KCl , 800 μM MgSO4 , 10 μM FeEDTA , 35 μM H3BO3 , 9 μM MnCl2 . 4H2O , 0 . 8 μM ZnCl2 , 0 . 5 μM Na2MoO4 . 2H2O , 0 . 3 μM CuSO4 . 5H2O , 6 mM KNO3 , 18 . 4 mM KH2PO4 , and 20 mM Na2HPO4 ) , and transferred to a controlled environment room ( 25°C , 16h light cycle ) . WT-lacZ and mutant SBW25 strains were grown overnight in M9 0 . 4% pyruvate media , then serially diluted in phosphate buffer . 1 x 103 CFU of mutant and WT-lacZ bacteria/plant were used to inoculate seven day-old wheat seedlings . Plants were grown for a further seven days , after which shoots were removed and 20 ml PBS was added to each tube and vortexed thoroughly to resuspend bacteria . A dilution series was plated onto XGal + IPTG + Cb plates and WT-lacZ/mutant colonies distinguished by blue/white selection . Assays were conducted for 10 plants/mutant , repeated at least twice independently , and statistical significance assessed using Mann-Whitney tests [43] . Bacterial growth was monitored in a microplate spectrophotometer ( BioTek Instruments ) with a minimum of 3 experimental replicates/sample . Wells ( of a 96-well plate ) contained 150 μL of the appropriate growth medium . Growth was initiated by the addition of 5μL of cell culture with an OD600 = 0 . 01 . Plates were covered with adhesive sealing sheets and incubated statically at 28°C or at 37°C for P . fluorescens and P . aeruginosa respectively . Each experiment was repeated at least twice independently . Chromatin Immunoprecipitation ( ChIP ) assays were performed as described elsewhere [44] , with the following modifications . After bacterial growth , formaldehyde ( 1% ) was added to the cultures and incubated at room temperature for 20 min , before the reaction was quenched with glycine ( 125 mM ) for 5 min . Cells were collected and washed with cold PBS four times . The cells were lysed in 1 ml of lysis buffer solution ( 10 mM Tris-HCl pH8 , 50 mM NaCl , 4 mg/ml lysozyme , 1x protease inhibitor ) . Cell extracts were then sonicated to fragment the DNA to an average size of 500bp . 50 μl of the extract was removed for total DNA preparation . For immunoprecipitation of RccR cross-linked DNA , a portion of the extract ( 1 ml ) was immunoprecipitated with 10 μl of polyclonal anti-RccR antibody at 4°C for 4h . After incubation with ProteinA affinity gel ( Sigma ) for 1h at 4°C , the beads were washed twice with IP buffer ( 100 mM Tris-HCl pH 8 . 0 , 250 mM NaCl , 0 . 5% Triton X-100 , 0 , 1% SDS , 1x protease inhibitor ) , and finally with TE buffer . The immunoprecipitated material was eluted with 100 μl of elution buffer ( 50 mM Tris-Cl pH 7 . 6 , 10 mM EDTA , 1% SDS ) . Cross-linking of immunoprecipitated and total DNA was reversed by incubation at 65°C overnight . After Proteinase K ( Roche ) treatment , the immunoprecipitate and total DNA samples were extracted using phenol-chloroform and purified using a QiaQuick kit ( Qiagen ) . Illumina TruSeq ChIP-seq libraries were produced from these samples and size selected to ~200–300 bp . , then sequenced on a single Illumina HiSeq 2000/2500 lane in High Output mode , using 100 bp single-end reads . The reads in the fastq files received from the sequencing contractor were aligned to the Pseudomonas fluorescens SBW25 genome using the bowtie2 software , which resulted in one SAM ( . sam ) file for each fastq file . All further operations described below were carried out using a combination of Perl scripts dependent on the BioPerl toolkit and R scripts . From each SAM file , coverage at ( i . e . number of reads mapping to ) each nucleotide position of the Pseudomonas fluorescens SBW25 genome was calculated and the output was saved in files , referred to here as coverage files . For each coverage file , a local enrichment was calculated in a moving window of 51 nucleotides ( nt ) moving in steps of 25 nucleotides as ( the sum of coverage at each nucleotide position in the 51-nt window ) divided by ( the sum of coverage at each nucleotide position in a 4 , 001-nucleotide window centred around the 51-nucleotide window ) . This results in an enrichment ratio value for every 25 nucleotides along the genome . All nucleotide positions where the enrichment ratio was less than 1 . 5 were removed , then a negative binomial distribution was fitted to the data using the fitdistr function of the MASS package in R . Thus , we arrived at the size and mu parameters of the binomial distribution . The values of size and mu parameters resulting from the fitting of the binomial distribution were then used to calculate p-values for each enrichment ratio using the pnbinom function of R . Finally , the p-values were adjusted for multiple testing by using the p . adjust function of R using the Benjamini and Hochberg method . This resulted in tables with three columns: Genomic position , Enrichment ratio and Adjusted P-value . Information on genes to the left and right of each genomic position was added to these tables using gene coordinates from the SBW25 GenBank file . The ChIP-seq data has been deposited in the ArrayExpress database ( accession number E-MTAB-5745 ) . β-galactosidase assays were performed as previously described [45] . Cells were grown to OD600 0 . 5–0 . 6 in M9 media supplemented with glucose/glycerol/pyruvate/acetate , then permeabilized with sodium dodecyl sulphate and chloroform . β-galactosidase activity was assayed using o-nitrophenyl-b-D-galactopyranoside ( ONPG ) as a substrate , and is reported as μmol o-nitrophenol/min/mg cellular protein . The results obtained were normalized to the UM recorded for WT strains carrying the plasmids pGm-p-rccR or pGM-p-hexR ( S4 Table ) . Experiments showing the expression of lacZ fusions in single-value data are the average of at least four independent measurements . Total RNA was extracted from 50ml cultures of SBW25/PA01 WT and ΔrccR/ΔhexR mutants grown in M9 minimal media to OD600 = 0 . 6 , or after Media exchange where appropriate ( below ) . 30 ml of 60% RNAlater ( in PBS ) was added to each tube , and sealed tubes were vortexed and centrifuged for 10 min at 4°C . The pellets were resuspended in 1x PBS + chilled β-mercaptoethanol RT solution , and lysed by mechanical disruption . Finally , RNA was purified from the lysate by column capture using an RNeasy Mini Kit ( Qiagen ) . Purified RNA was subjected to an additional DNase treatment ( Turbo™ DNase , Ambion ) . RNA quantification was performed with an ND-1000 Spectrophotometer . WT and mutant strains were grown in LB overnight at 28°C . The next day , the overnight cultures were diluted in M9 pyruvate and grown until OD600 = 0 . 3 . Cells were then pelleted by centrifugation and resuspended in an equal volume of Rooting Solution without any carbon sources ( see Colonisation assay for composition ) . After 1 hour of incubation at 28°C , the cells were pelleted by centrifugation , then resuspended in 50 ml M9 supplemented with 0 . 4% glucose/ glycerol/ pyruvate/ acetate . Cultures were incubated for 30 min at 28°C , and RNA was extracted as described above . cDNA synthesis was performed as previously described [4] . Real time PCR was performed using a 20 μl reaction mix containing 1 μl cDNA . Primers from number 29 to 75 listed in S5 Table were used for the tested gene expressions . At least three wells were run for each sample . The amount of gene transcript in each case was analysed by Absolute or Relative studies ( 2−ΔΔCt method ) [46 , 47] . Absolute quantification was used to determine the number of copies of gene targets ( rccR or hexR ) in our WT strains without reference to other samples . For this analysis a standard curve was constructed ( in duplicate ) using SBW25 chromosomal DNA for the quantification of transcripts . All gene quantifications were normalized to the levels of the endogenous gene rpoD . Relative quantification was used to compare the abundance of rccR ( or hexR ) gene target mRNAs in equivalent WT and ΔrccR ( or ΔhexR ) samples . The amount of each gene transcript was normalized to the WT reference sample . For the 2−ΔΔCt method , results were presented as n-fold increase relative to the reference sample . The ΔCt-values were examined using the Student's t test to determine whether datasets for relative gene expression were significantly different from those in a chosen calibrator . Primers for the aforementioned transcripts and for the rpoD transcript , used as an endogenous control , were experimentally validated for suitability to the 2−ΔΔCt method , and are listed in S5 Table . Melting curve analysis was used to confirm the production of a specific single product from each primer pair . Each experiment was repeated at least twice independently . The rccR gene was cloned into pET42b to give the expression vector pET42b-rccR . This construct was transformed into E . coli strain BL21 ( DE3 ) . 25 ml of an overnight culture of BL21pLys pET42b-rccR were used to inoculate 4L of LB supplemented with Kc . Cells were grown at 28°C to an OD600 of 0 . 4 , then 1 mM isopropyl β-D-thiogalactopyranoside was added to induce rccR expression . After 2h shaking , the cells were harvested by centrifugation , the pellet was resuspended in 40ml Equilibration buffer ( 20mM Hepes , 250mM NaCl , 10mM MgCl2 , 2 . 5% glycerol , pH 6 . 8 ) and lysed by sonication . An additional centrifugation step removed cell debris , then the supernatant was applied to 1 ml nickel-charged HisTrap chelating column attached to an Äkta FPLC ( GEHealthcare ) . The column was washed with Equilibration buffer , while the protein was eluted with a linear gradient of imidazole to 1M in the Elution Buffer ( 20mM Hepes , 150mM NaCl , pH7 . 4 ) . The fractions containing RccR-His were pooled and applied to 5ml heparin column . The column was washed with Buffer A ( 20mM Hepes , 50mM NaTiocianate , pH7 . 4 ) , while the elution was conducted using Buffer B ( 20mM Hepes 1M NaCl , pH7 . 4 ) . The fractions containing RccR were then combined . The C-terminal His tag was not removed from the purified protein for any of the subsequent experiments . Transcriptional start site identification for the RccR gene targets was carried out using the Invitrogen kit 5’ RACE System for Rapid Amplification of cDNA Ends , Version 2 . 0 . Total RNA were purified from P . fluorescens as above , then mRNA enrichment was carried out using a specific primer GSP1 for the RNA target . After reverse transcription to cDNA using SuperScriptII RT , RNA was degraded from the samples using an RNase mix . The cDNA was purified using GlassMAX Spin Cartridge , then tailed with dCTP and TdT . After the tailing reaction , a second PCR reaction using the Abridged Anchor Primer ( AAP ) and a GSP2 nested primer was performed . Finally , a third PCR to re-amplify the previous PCR product was conducted , using the AUAP primer ( that recognises the AAP region ) , and a GSP3 nested primer . The final PCR product was sequenced with Big Dye 3 . 1 by Eurofins . All the SPR experiments described here were performed using a GE Healthcare Biacore T200 instrument . All measurements were recorded at 25°C using the ReDCaT system described in [32] , with a single SensorChip SA ( GE Healthcare ) having four flow cells each containing SA pre-immobilized to a carboxymethylated dextran matrix . DNA samples were purchased from Eurofins Genomics as desalted single-stranded ( ss ) oligomers at 100 μM concentration in water , while the RccR-His protein was purified as described above . These experiments were performed based on knowledge of the predicted RccR consensus sequences , so the DNA fragments were designed ad hoc , with a length of about 30 nt each ( S5 Table , 93/110 ) . The DNA was prepared by taking 45 μl of the shorter strand and mixing with 55 μl of the longer strand . This was then heated to 95°C and cooled to room temperature . This gave a 45 μM stock which was then diluted to 1 μM in running buffer ( 10 mM Hepes pH 7 . 4 , 300 mM NaCl , 3mM EDTA , 0 . 05% v/v tween20 ) . For each experiment , all protein samples were diluted in running buffer . For all the SPR experiments the DNA at 1uM was captured at the start of each cycle by loading at 10μl/min on FC2 . Interaction with RccR protein was measured by flowing the RccR protein over both FC1 and FC2 for 60s at a flow of 30μl/min . The chip was then washed to remove any protein still bound and the DNA by washing with 1 M NaCl , 50mM NaOH at the end of each cycle . Initial screening experiments were run to measure RccR binding to each DNA sequence target . Concentrations of 1 μM and 0 . 1 μM of RccR protein were used . The amount of DNA captured and the binding response of RccR was measured and this was converted to the %Rmax bound ( percentage of the theoretical maximum response ) , assuming a single RccR dimer binding to a single immobilized ds DNA oligomer ( S6 Table ) . The affinity of RccR for the aceE , aceA and rccR consensus sequences was then determined , The RccR protein stock ( 40 mM ) was serially diluted in running buffer to give concentrations of 100 , 50 , 25 , 12 . 5 , 6 . 25 , 3 . 125 , 1 . 56 , 0 . 78 , 0 . 39 and 0 . 19 μM . Measurements were taken in triplicate for a range of protein concentrations spanning either side of the expected Kd using a multicycle kinetic approach . For RccR ligand identification , binding experiments were run to measure the association of RccR with aceE , aceA and rccR consensus sequences in the presence of different carbon metabolites . Several samples ( RccR—hypothetical ligand mixes ) were prepared , so that RccR protein was diluted to 0 . 5 μM in running buffer containing the tested ligand at different concentrations ( 0-1-10-100-1000 μM ) with a final composition matching the running buffer . The %Rmax bound , assuming a single RccR dimer binding to one immobilized ds DNA oligomer was calculated ( S6 Table ) . The regions of the P . fluorescens SBW25 genome containing the predicted RccR binding sites upstream of the coding sequences of aceA , aceE and rccR were PCR amplified with oligonucleotides 81–82 , 83–84 , 85–86 , respectively , and the generated DNA fragments inserted into the plasmid pGEM-T-Easy ( Promega ) . The DNA probes were 5’-end labelled and the assays carried out as previously described [48] . Briefly , 1 pmol of pGEM-T-Easy-PaceA , pGEM-T-Easy-PaceE and pGEM-T-Easy-PrccR were linearized by SpeI digestion , dephosphorylated with calf intestinal phosphatase and 5’-end labelled with T4 polynucleotide kinase in the presence of 2 pmol of [γ32-P]-ATP ( 6000 Ci/mmol; Perkin Elmer ) . Following NcoI digestion , the labelled DNA fragments were gel purified and resuspended in ddH2O . For footprinting assays , approximately 20 fmol of the labelled probes were incubated with purified RccR protein in 50 μL of 1X Footprinting Buffer ( 10 mM Tris-HCl pH 8 . 0; 50 mM NaCl; 10 mM KCl; 5 mM MgCl2; 1 mM DTT; 0 . 01% NP-40; 10% glycerol ) containing 200 ng of sonicated salmon sperm DNA as a non-specific competitor , for 15 min at room temperature; afterwards , 0 . 04 units of DNase I ( EMD Millipore ) , freshly diluted in Footprinting Buffer containing 5 mM CaCl2 , were added to the reaction mixture , and the digestion was allowed to occur for 75 seconds at room temperature . After stopping the reactions , samples were phenol-chloroform extracted , ethanol precipitated and resuspended in Formamide Loading Buffer ( 95% formamide; 10 mM EDTA; 0 . 02% bromophenol blue; 0 . 02% xylene cyanol ) . Next , samples were denatured at 100°C for 5 min , separated on 8 M urea–6% polyacrylamide sequencing gels in TBE buffer and autoradiographed . Genomic regions P . fluorescens SBW25 encompassing RccR binding sites on aceE , aceA and rccR promoters were PCR amplified with specific primers ( 87–88 , 89–90 , 91–92 ) and cloned into the plasmid pGEM-T-Easy . Then , DNA probes were 5'-end labelled ( using BamHI and XhoI for the sequential digestions ) and purified as described above for DNase I footprintings . Hydroxyl-radical footprinting experiments were performed as previously described [49] with some modifications . Approximately 20 fmol of labelled probes were incubated with increasing concentrations of RccR protein in OH-Footprinting Buffer ( 10 mM Tris-HCl , pH 8 . 0; 50 mM NaCl; 10 mM KCl; 5 mM MgCl2; 0 . 1 mM DTT; 0 . 01% NP40 ) for 15 minutes at room temperature , including 200 ng of sonicated salmon sperm DNA as a non-specific competitor in a final volume of 30 μL . The digestions of the labelled probes were performed using 2 μL each of the following solutions: 125 mM Fe ( NH4 ) 2 ( SO4 ) 2- , 250 mM EDTA , 1% H2O2 and 100 mM DTT . After 2 minutes , for each sample the reaction was quenched by the addition of 25 μL of OH Stop Buffer ( 4% glycerol; 600 mM NaOAc , pH 5 . 2; 100 ng/μL sonicated salmon sperm DNA ) . Samples were phenol/chloroform extracted , ethanol precipitated and resuspended in 12 μL of Formamide Loading Buffer . Next , samples were denatured at 100°C for 5 minutes , separated on an 8M urea-8 . 5% polyacrylamide sequencing gel in TBE buffer and autoradiographed .
|
Here we show how Pseudomonas controls multiple different primary carbon metabolism pathways by sensing levels of KDPG , an Entner Doudoroff ( ED ) pathway intermediate . KDPG binds to two highly similar transcription factors; the ED regulator HexR and the previously uncharacterised protein RccR . RccR inversely controls the glyoxylate shunt , gluconeogenesis and pyruvate metabolism , suppressing the first two pathways as pyruvate metabolism genes are expressed , and vice versa . This complex regulation is enabled by two distinct RccR-binding consensus sequences in the RccR regulon promoters . KDPG binding simultaneously increases RccR affinity for the glyoxylate shunt and gluconeogenesis promoters , and releases repression of pyruvate metabolism . This elegant two-regulator circuit allows Pseudomonas to rapidly respond to carbon source availability by sensing a single key intermediate , KDPG .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"carbohydrate",
"metabolism",
"protein",
"metabolism",
"chemical",
"compounds",
"ketones",
"glucose",
"metabolism",
"pyruvate",
"regulator",
"genes",
"monomers",
"(chemistry)",
"gene",
"types",
"sequence",
"motif",
"analysis",
"bacteria",
"research",
"and",
"analysis",
"methods",
"pseudomonas",
"sequence",
"analysis",
"polymer",
"chemistry",
"bioinformatics",
"acids",
"gene",
"expression",
"chemistry",
"biochemistry",
"pseudomonas",
"fluorescens",
"database",
"and",
"informatics",
"methods",
"glycerol",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"metabolism",
"organisms"
] |
2017
|
One ligand, two regulators and three binding sites: How KDPG controls primary carbon metabolism in Pseudomonas
|
In response to insect attack and mechanical wounding , plants activate the expression of genes involved in various defense-related processes . A fascinating feature of these inducible defenses is their occurrence both locally at the wounding site and systemically in undamaged leaves throughout the plant . Wound-inducible proteinase inhibitors ( PIs ) in tomato ( Solanum lycopersicum ) provide an attractive model to understand the signal transduction events leading from localized injury to the systemic expression of defense-related genes . Among the identified intercellular molecules in regulating systemic wound response of tomato are the peptide signal systemin and the oxylipin signal jasmonic acid ( JA ) . The systemin/JA signaling pathway provides a unique opportunity to investigate , in a single experimental system , the mechanism by which peptide and oxylipin signals interact to coordinate plant systemic immunity . Here we describe the characterization of the tomato suppressor of prosystemin-mediated responses8 ( spr8 ) mutant , which was isolated as a suppressor of ( pro ) systemin-mediated signaling . spr8 plants exhibit a series of JA-dependent immune deficiencies , including the inability to express wound-responsive genes , abnormal development of glandular trichomes , and severely compromised resistance to cotton bollworm ( Helicoverpa armigera ) and Botrytis cinerea . Map-based cloning studies demonstrate that the spr8 mutant phenotype results from a point mutation in the catalytic domain of TomLoxD , a chloroplast-localized lipoxygenase involved in JA biosynthesis . We present evidence that overexpression of TomLoxD leads to elevated wound-induced JA biosynthesis , increased expression of wound-responsive genes and , therefore , enhanced resistance to insect herbivory attack and necrotrophic pathogen infection . These results indicate that TomLoxD is involved in wound-induced JA biosynthesis and highlight the application potential of this gene for crop protection against insects and pathogens .
Higher plants respond to insect attack and wounding by activating the expression of genes involved in herbivore deterrence , wound healing , and other defense-related processes [1]–[7] . The wound response of tomato ( Solanum lycopersicum ) provides an attractive model to understand the signal transduction events leading from localized injury to the systemic expression of defense-related genes [7] , [8] . The principle defensive markers used in these studies are genes encoding proteinase inhibitors ( PIs ) , low molecular weight proteins that inhibit the activity of digestive enzymes in the gut of herbivores [1] , [9] . In their milestone study of wound-inducible PIs in tomato , Green and Ryan proposed that specific signals generated at the wound site travel through the plant and activate the expression of PIs and other defense-related genes in remote responding leaves [10] . Systemin , an 18-amino-acid peptide signal , was purified from wounded tomato leaves on the basis of its ability to activate PI accumulation using a convenient bioassay for PI-inducing compounds [9] , [11]–[13] . Systemin is derived from the cleavage of a larger precursor protein called prosystemin , which is encoded by a single copy of the Prosystemin ( PS ) gene [12] , [14] . Transgenic tomato plants that express an antisense PS are defective in wound-induced systemic expression of PI genes and are more susceptible to insects [14] . Conversely , transgenic tomato plants ( called 35S::PS ) that overexpress PS constitutively express high levels of PIs without wounding and are more resistant to insects [15] , [16] . In addition , genetic analysis in tomato has shown that genes required for ( pro ) systemin signaling are also essential for wound-induced expression of defensive genes [3] , [17] , [18] . Together , these genetic studies support that the peptide signal systemin acts as an upstream component of the wound-induced signaling cascades leading to defense gene expression . It is generally believed that wounding and insect attack lead to the rapid cleavage of systemin from prosystemin . Binding of systemin to its proposed receptor on the cell surface then activates defense gene expression by increasing the endogenous levels of jasmonic acid ( JA ) and related pentacyclic oxylipins ( collectively referred to here as JAs ) that are derived from the linolenic acid via the octadecanoid pathway [1] , [19]–[21] . A role for JAs in intercellular signaling is supported by the fact that application of MeJA ( the methyl ester of JA ) to one tomato leaf induces PI expression in distal untreated leaves [22] . JAs are now considered to be key regulators for stress-induced gene expression in virtually all plant species [1] , [20] , [23]–[27] . It was proposed that systemin and JA work together in the same signal transduction pathway to regulate the systemic expression of defense-related genes [1] , [9] , [20] . Thus , the systemin/JA signaling pathway for induced resistance in tomato provides a unique opportunity to investigate , in a single experimental system , the mechanism by which peptide and oxylipin signals interact to coordinate systemic expression of defense-related genes [7] , [8] . We have been using a genetic approach to dissect the systemin/JA signaling pathway and to elucidate the role of systemin and JA in it . Genetic screen to identify mutations that suppress the constant wound signaling phenotype ( i . e . , constitutive expression of PIs and other defense-related genes ) of 35S::PS plants has led to the identification of several important components of the systemin/JA signaling pathway 17 , 18 , 28 , 29 . Significantly , several of the characterized spr ( suppressors of prosystemin-mediated responses ) mutants actually define genes that are directly involved in JA biosynthesis or signaling [17] , [18] , [29] . For example , Spr2 encodes a chloroplast fatty acid desaturase that catalyzes the ω3 desaturation of linoleic acid ( 18∶2 ) to linolenic acid ( 18∶3 ) , the metabolic precursor for JA biosynthesis [18] . spr6 , on the other hand , defines the tomato homolog of CORONATINE INSENSITIVE1 ( COI1 ) , which has been shown to be the JA receptor in Arabidopsis [29] , [30] . These studies provided direct evidence that JA acts downstream of systemin in regulating wound-induced expression of defense-related genes . Grafting experiments conducted with the JA biosynthesis mutant spr2 and the JA signaling mutant jai-1 revealed that systemic defense signaling requires both the biosynthesis of JA at the site of wounding and the ability to perceive JA in remote tissues , suggesting that JA acts as a systemic wound signal [3] . Grafting experiments also demonstrated that the graft-transmissible wound signal generated by the 35S::PS plants can be readily recognized by spr2 plants ( which are insensitive to systemin ) , but cannot be recognized by the JA-insensitive jai-1 plants , strongly suggesting that the 35S::PS-derived wound signal is JA , rather than systemin [3] . These results challenge the previous paradigm that systemin is the long-distance mobile signal for wound-induced defense gene expression [8] , [31] , [32] . Genetic analyses of tomato wound response also provide insight to understand how the peptide signal systemin interacts with JA to promote systemic defense signaling . In contrast to other tomato wound response mutants that lack both local and systemic PI expression in response to wounding , spr1 plants were deficient mainly in the systemic response . Moreover , spr1 abolished JA accumulation in response to exogenous systemin , and showed reduced JA accumulation in wounded leaves [28] Analysis of reciprocal grafts between spr1 and wild-type ( WT ) plants showed that spr1 impedes systemic PI expression by blocking the production of the long-distance wound signal in damaged leaves , rather than inhibiting the recognition of that signal in systemic undamaged leaves . These experiments suggest that Spr1 is involved in a signaling step that couples systemin perception to the activation of the octadecanoid pathway [28] . These and other studies support that systemin acts locally at the site of wounding to amplify the production of JA , which in turn functions as a mobile signal to activate systemic defense responses [8] , [28] , [33] . In addition to systemin , the hydroxyproline-rich glycopeptides ( HypSys peptides ) , which are isolated from tomato and tobacco leaves , are also powerful activators of PI expression [34] . Recent genetic data support that , similar to systemin , HypSys peptides also play a role in an amplification loop that upregulates JA production to effect strong systemic defense response [35] . Toward understanding the molecular mechanism of systemin/JA-mediated systemic defense signaling in tomato , we are conducting an enlarged genetic screen to identify more spr mutants that suppress the constitutive wound signaling phenotype of the 35S::PS plants [29] . Here we report the genetic and molecular characterization of spr8 , a semidominant mutant that is defective in wound-induced expression of defense-related genes . Map-based cloning studies reveal that Spr8 encodes tomato lipoxygenase D ( TomLoxD ) , a 13-lipoxygenase that catalyzes the hydroperoxidation of linolenic acid , a key step in JA biosynthesis [19] . We show that overexpression of TomLoxD leads to elevated wound-induced JA biosynthesis , increased expression of wound-responsive genes and , therefore , enhanced resistance to insects and necrotrophic pathogens . These results highlight the application potential of the TomLoxD gene for crop protection .
spr8 is one of the newly identified mutants that can block the constitutively high activity of polyphenol oxidase ( PPO ) in the 35S::PS plants [29] . Further characterization of spr8 was carried out using a spr8/spr8 homozygous line in which the 35S::PS transgene was removed by five successive backcrosses to the WT cv . Castlemart ( CM ) . The overall plant morphology , flower development and pollen viability of spr8 plants were indistinguishable from those of WT plants ( Figure S1 ) . The wound response of spr8 was compared with that of WT using the classical radial immunodiffusion assay for the measurement of wound-induced accumulation of proteinase inhibitor II ( PI-II ) [11] , [29] , [36] . For these experiments , 16-day-old seedlings containing two fully expanded leaves were wounded and the accumulation of PI-II protein was quantified . Wounding the lower leaves of WT caused the well-known accumulation of PI-II both in the wounded leaves ( local response ) and in the upper unwounded leaves ( systemic response ) ( Figure 1A ) . In contrast , spr8 seedlings accumulated no detectable PI-II protein in the wounded leaves and the upper unwounded leaves ( Figure 1A ) . Consistent with the PI-II protein data , quantitative real-time PCR ( qRT-PCR ) assays indicated that the PI-II transcripts were expressed very weakly in wounded spr8 leaves as compared to those in WT leaves ( Figure 1B ) . It has been shown that , similar to the PI genes [37] , protein products of the tomato wound-responsive genes threonine deaminase ( TD ) [16] and leucine amino peptidase A ( LapA ) [38] have a direct role in deterring insect performance . Our parallel experiments indicated that the wound-induced expression levels of TD ( Figure S2A ) and LapA ( Figure S2B ) were also largely reduced in spr8 plants compared to those in WT plants . These results demonstrate that the spr8 mutation impairs wound-induced expression of defensive genes . To gain additional insight into the wound response phenotype of spr8 , we examined the capacity of the mutant to respond to various PI-inducing compounds . As previously reported [28] , exogenous application of systemin led to strong expression of PI-II transcripts in WT plants ( Figure 1C ) . But spr8 plants failed to express significant levels of PI-II transcripts in response to the same concentrations of systemin ( Figure 1C ) , indicating that spr8 plants are insensitive to systemin . These results are consistent with the fact that spr8 was identified as a suppressor of prosystemin-mediated responses . We then examined the response of spr8 to the methyl ester of JA , MeJA , which is a potent elicitor of PI-II expression in WT plants ( Figure 1D ) . As shown in Figure 1D , exogenous application of MeJA readily restored the PI-II expression of spr8 mutants to levels comparable to those of WT plants . These results led us to classify spr8 into the group of wounding/systemin-insensitive , but JA-sensitive mutants . It is most likely that the spr8 mutant defines a signaling step that couples the perception of systemin to activation of the JA pathway . Trichome density and volatile emissions of glandular trichomes provide a formidable protective barrier to invasion by herbivores and pathogens [39]–[41] . Cultivated tomato contains two morphologically distinct types of glandular trichomes . Type I trichomes have an elongated multicellular stalk with a small unicellular vesicle at the tip ( Figure 2A and 2B ) . Type VI trichomes have a unicellular stalk with a four-celled glandular head ( Figure 2A and 2B ) [42] , [43] . In order to determine whether spr8 affects trichome development , we used scanning electron microscopy to observe the adaxial leaf surface to compare trichome morphology and density between WT and spr8 plants . A striking feature of spr8 leaves is the significant reduction of trichome number of both types ( Figure 2A and 2B ) . Quantification of trichomes of five-week-old WT plants ( containing at least five leaves ) showed that the density of type VI trichome was ∼10 trichomes/mm2 on the base region of the third leaflet . Analysis of comparable spr8 leaflets showed that , type VI trichome density of the mutant was about 70% of that of WT leaflets ( Figure 2C ) . Next , we used gas chromatography analysis to determine whether spr8 affects the production of compounds that are synthesized in trichome glands . For these experiments , type VI glandular trichomes were selectively collected by using a stretched-glass pipette and were extracted with methyl tert-butyl ether ( MTBE ) ( see method ) . Trichome exudates were then analyzed by gas chromatography to measure the terpene composition . From 1 , 000 type VI glands collected from the adaxial surface of WT leaves , six monoterpenes ( α-pinene , β-myrecene , 2-carene , α-phellandrene , β-phellandrene and limonene; Figure 2F ) and three sesquiterpenes ( δ-elemene , β-caryophyllene , and α-humulene; Figure 2G ) were identified . Comparison of terpene profiles revealed that , all of these compounds were also detected in exudates from the same number of type VI glandular trichomes of spr8 leaflets , but their accumulation levels were significantly decreased in the mutant ( Figure 2F and 2G ) . In spr8 glandular trichomes , the accumulation levels of total monoterpenes and sesquiterpenes were reduced to 19 . 5% and 15 . 2% , respectively , of those of their WT counterparts ( Figure 2D and 2E ) . These results support the hypothesis that the spr8 mutation affects the terpene metabolic pathway that mainly operates in type VI trichome glands . The inability of spr8 plants to express significant levels of defensive genes in response to mechanical wounding and systemin ( Figure 1 and Figure S2 ) suggests that this mutant is compromised in resistance to herbivorous insects . To test this hypothesis , newly hatched cotton bollworm ( Helicoverpa armigera ) larvae were placed on leaves of 5-week-old plants to initiate a feeding trial . Sustaining long-term feeding by insects , spr8 plants were severely damaged ( Figure 3A , right ) , while WT plants showed relatively few signs of macroscopic damage during the period of the feeding trial ( Figure 3A , left ) . After termination of the feeding trial , PI-II protein accumulation in the remaining leaf tissues was measured , as was the weight gain of larvae reared on both of the host genotypes . In contrast with high levels of PI-II accumulation in herbivore-damaged WT leaves , very little or no PI-II protein accumulation was detected in hornworm-challenged spr8 plants ( Figure 3B ) . These results indicate that WT plants have relatively high levels of natural resistance to the cotton bollworm larvae and that this resistance is severely compromised by the spr8 mutation . Consistently , the average weight of larvae reared on the mutant was 2 . 0-fold greater than that of larvae reared on WT plants ( Figure 3C and 3D ) . These results demonstrate that Spr8 is required for the resistance of tomato plants to attacking hornworm larvae . Genetic analysis revealed that spr8 is a semi-dominant mutant , given that the wound-response phenotype of the heterozygous ( Spr8/spr8 ) plants was intermediate between those of the homozygous spr8 plants and their WT counterparts ( Figure 1A and Figure S3 ) . The deficiency in wound-induced PI-II protein accumulation of spr8 provides a facile assay for map-based cloning studies to determine the genetic basis of this defect . A combination of cleaved amplified polymorphic sequence ( CAPS ) and simple sequence repeat ( SSR ) markers was used to localize Spr8 to a region on the long arm of chromosome 3 between SSR markers TES0023 and TES1203 ( Figure 4A ) . Fine mapping using 354 backcrossed ( BC1 ) individuals showing a WT wound response delimited the Spr8 locus to a region between the markers SSR601 and M140 in the scaffold SL2 . 40sc03701 of the sequenced tomato genome [44] , [45] . Among the genes predicted by the International Tomato Annotation Group ( ITAG2 . 3 release , http://solgenomics . net ) in this interval , Solyc03g122340 , which encodes TomLoxD ( tomato lipoxygenase D ) , a wound-inducible lipoxygenase [46] , is considered to be a strong candidate of Spr8 . DNA sequencing revealed that spr8-derived TomLoxD complementary DNA ( cDNA ) contains a single C-to-T mutation ( Figure 4B ) . This C-to-T mutation , which was confirmed by sequencing of PCR-amplified genomic DNA from spr8 plants , destroys a BamHI restriction site , and a CAPS marker was developed to detect the spr8 mutant allele ( Figure 4C ) . The single base pair change in the TomLoxD gene is predicted to replace a highly conserved ( i . e . , invariant among plant and animal lipoxygenases ) Pro residue at position 598 with an Leu ( Figure 4D and Figure S4 ) . Considering that spr8 is a semi-dominant mutation , we performed the following experiments to show that the missense mutation in TomLoxD accounts for the wound response phenotype of spr8 . First , transgenic plants overexpressing a WT allele of TomLoxD ( TomLoxD-OE ) showed increased wound response in term of wound-induced defense gene expression ( See below ) . Second , similar to spr8 plants , transgenic plants expressing a TomLoxD RNA interference ( RNAi ) construct ( TomLoxD-RNAi ) were defective in wound-induced expression of PI-II ( Figure S5A and S5B ) . Third , the wound response phenotype of transgenic plants overexpressing a mutant allele of TomLoxD ( TomLoxDP598L-OE ) was intermediate between that of the homozygous spr8 plants and their WT counterparts ( Figure S5A and S5B ) . Finally , overexpression of a WT allele of TomLoxD in the spr8 background failed to fully rescue the wound response defects of the mutant ( Figure S5C and S5D ) . Collectively , these results support that the identified C-to-T mutation in the TomLoxD gene is responsible for the wound response phenotype of spr8 plants and that the spr8 allele of TomLoxD ( i . e . , TomLoxDP598L ) acts as a dominant negative regulator of the tomato wound response pathway . Lipoxygenases are nonheme iron-containing fatty acid dioxygenases that catalyze the peroxidation of polyunsaturated fatty acids such as linoleic acid , α-linolenic acid , and arachidonic acid [47] . Based on the positional specificity of linoleic acid oxygenation , they are classified as 9-lipoxygenases ( oxygenation occurs at carbon 9 of the hydrocarbon backbone ) and 13-lipoxygenases ( oxygenation occurs at carbon 13 of the hydrocarbon backbone ) . 13-lipoxygenases can be further divided as types 1 and 2 based on the presence of a putative chloroplast transit peptide ( cTP ) [47] . ChloroP ( http://www . cbs . dtu . dk/services/TargetP/ ) -based analysis predicted that the deduced amino acid sequence of TomLoxD contains a putative cTP ( TomLoxD1–77 ) , a small N-terminal PLAT/LH2 domain ( TomLoxD78–213 ) that forms a β-barrel , and a C-terminal domain ( TomLoxD222–892 ) that forms α-helices ( Figure 4D ) . It is generally believed that the N-terminal β-barrel domain is involved in membrane or substrate binding , whereas the C-terminal domain harbors the catalytic site of the enzyme [48] . This primary protein structure suggests that TomLoxD is a member of the type 2 plastid-localized 13-lipoxygenases [47] . This prediction is supported by our phylogenetic analysis of plant lipoxygenases , which places TomLoxD in a clade including functionally characterized and predicted type 2 13-lipoxygenases ( Figure 4E ) . To confirm the chloroplast localization of the TomLoxD protein , full-length of the TomLoxD cDNA was fused to the green fluorescent protein ( GFP ) reporter gene and subsequently transformed into Arabidopsis leaf protoplast cells . As shown in Figure S6 , the GFP fluorescence was co-localized with the red chlorophyll autofluorescence , suggesting that TomLoxD is a chloroplast-localized protein . Notably , in our phylogenetic analysis , TomLoxD was most similar to the Arabidopsis LOX3 and LOX4 ( 71 . 7% and 71 . 3% amino acid identity , respectively ) ( Figure 4E ) , which has recently been shown to be type 2 chloroplast-localized 13-lipoxygenases that are involved in JA biosynthesis [49] . It is noteworthy that the TomLoxDP598L mutation in spr8 occurs in the C-terminal α-helices domain , presumably impairs the catalytic activity of the enzyme ( Figure 4D ) . The above-described results point to a possibility that TomLoxD is a functional 13-lipoxygenase involved in wound-induced JA biosynthesis and that the spr8 allele of TomLoxD ( hereafter referred to as TomLoxDP598L ) impairs wound-induced JA biosynthesis . As the first step to prove this , we examined the expression of TomLoxD or TomLoxDP598L in response to wounding . Consistent with a previous investigation [46] , the levels of TomLoxD transcripts were induced by wounding within 30 min and peaked at 1 h after wounding , TomLoxD transcripts then showed a tendency of decline and returned to control levels within 8 h ( Figure 5A ) , indicating that TomLoxD is an early wound-inducible gene . Interestingly , the wound-induced expression kinetics of TomLoxDP598L was essentially similar to that of TomLoxD , albeit its expression levels were somehow reduced as compared to that of the latter ( Figure 5A ) . These results indicate that TomLoxDP598L is still responsive to wounding . To determine the contribution of TomLoxD and TomLoxDP598L in wound-induced JA biosynthesis , we used liquid chromatography coupled with tandem mass spectrometry ( LC-MS/MS ) to measure endogenous JA levels in WT and spr8 plants in response to wounding . We consistently observed that the JA levels in unwounded WT and mutant leaves were below the detection limit ( Figure 5B ) . One hour after wounding , the average JA level was increased to 31 . 7±1 . 1 pg per milligram of fresh weight ( pg/mg FW ) in WT leaves , whereas the average JA level in mutant leaves was only 7 . 9±0 . 3 pg/mg FW ( P<0 . 0001 , Student's t test ) ( Figure 5B ) , confirming that spr8 plants are defective in wound-induced JA biosynthesis . These results indicate that TomLoxD is required for wound-induced JA biosynthesis and that the TomLoxDP598L mutant allele largely impairs this capability . Taken together , our data support that , even though the expression of TomLoxDP598L is still responsive to mechanical wounding ( Figure 5A ) , this mutant version of TomLoxD impairs wound-induced JA biosynthesis ( Figure 5B ) . In the model plant of Arabidopsis , much of our understanding of the JA signaling has come from the recent elucidation of the molecular details of JA-regulated gene transcription through the basic helix-loop-helix ( bHLH ) -type transcription factor MYC2 , a master regulator of JA responses [50]–[54] . Considering that in Arabidopsis MYC2 directly regulates the expression of several JA biosynthetic genes including LOX2 [55] , it is reasonable to speculate that SlMYC2 , the tomato homolog of MYC2 , may directly regulate the expression of TomLoxD . Indeed , several lines of evidence lends support to this hypothesis . First , wound-induced expression levels of TomLoxD were substantially reduced in SlMYC2-RNAi plants as compared to those in WT plants ( Figure 6A , 6B and Figure S7 ) , indicating that SlMYC2 positively regulates the wound-induced expression of TomLoxD; Second , chromatin immunoprecipitation ( ChIP ) assays using 35Spro:SlMYC2-4myc plants indicated that SlMYC2 associates with a G-box-like motif ( CCATGTG ) in the promoter region of TomLoxD ( Figure 6C and 6D ) ; Third , DNA electrophoretic mobility shift assays ( EMSA ) indicated that a maltose binding protein ( MBP ) -SlMYC2 fusion protein binds the promoter of TomLoxD in a G-box-like motif-dependent manner ( Figure 6E ) . Finally , using the transient expression assay of Nicotiana benthamiana leaves , we verified the activation effect of SlMYC2 on the expression of a reporter containing the TomLoxD promoter fused with the firefly luciferase gene ( LUC ) ( Figure 6F and 6G ) . Together , these data demonstrate that the wound-induced expression of TomLoxD is under the direct regulation of SlMYC2 . Our findings that TomLoxD is required for wound-induced JA biosynthesis and defense gene expression raised the possibility that overexpression of this gene could enhance wound-induced JA biosynthesis , which , in turn , leads to increased plant resistance . To test this hypothesis , we generated transgenic tomato plants overexpressing the TomLoxD cDNA driven by the cauliflower mosaic virus 35S promoter ( OE plants ) . Increased expression of TomLoxD in transgenic lines including OE-1 , OE-3 and OE-5 was confirmed by qRT-PCR analysis ( Figure 7A ) . Under normal growth conditions , the overall growth and morphology of these OE plants was essentially similar to those of WT plants ( Figure S1 ) . We then compared the expression levels of defensive genes between these OE plants and WT plants . Similar steady-state levels of PI-II , TD and LapA transcripts were detected between the noninduced OE plants and WT plants ( Figure 7B–7D ) . A marked increase in the accumulation levels of these transcripts was , however , observed in the TomLoxD overexpression plants in response to mechanical wounding ( Figure 7B–7D ) . These results demonstrate that overexpression of TomLoxD leads to enhanced wound-induced activation of PI-II and other defense-related genes . To test that the increased wound-induced defense gene expression in these OE lines may be resulted from enhanced wound-induced accumulation levels of JA , we examined wound-induced JA accumulation between OE-5 and WT plants . Similar steady-state levels of JA were detected between OE-5 and WT plants ( Figure 7E ) , indicating that overexpression of TomLoxD does not lead to constant accumulation of high levels of JA . In response to mechanical wounding , however , a substantial increase in the accumulation of JA was observed in OE-5 plants ( Figure 7E ) , indicating that overexpression of TomLoxD leads to enhanced wound-induced accumulation of the defense hormone JA . The ability of TomLoxD overexpresser lines to accumulate higher levels of JA and to express increased levels of defensive genes in response to mechanical wounding suggested that these transgenic plants may be more resistant to herbivorous insects . To test this possibility , five-week-old OE-5 and WT plants were challenged with Helicoverpa armigera larvae . After termination of the feeding trial , we examined the weight of the larvae to assess the resistance of plants . The average weight of larvae reared on OE-5 plants was only 32 . 5% of that of larvae reared on WT plants ( Figure 6F–6H ) , demonstrating that overexpression of TomLoxD leads to enhanced plant resistance to herbivorous insects . Considering that the JA-signaled plant resistance is also effective to the necrotrophic pathogen Botrytis cinerea [50] , [51] , [56]–[58] , we examined the performance of OE-5 plants to the Hy2-1 strain of B . cinerea . For these experiments , detached leaves from five-week-old tomato plants were inoculated with 5 µL 5×105 per mL spore suspension and disease development was analyzed 3 days after inoculation ( DAI ) . As measured by the size of necrotic lesions , whereas spr8 plants were more susceptible than WT plants to B . cinerea infection , OE-5 plants were more resistant than WT plants to this pathogen ( Figure 7I and 7J ) . In another pathogen infection assay , 16-day-old seedlings were inoculated in planta with spore suspensions of B . cinerea and the expression levels of the pathogenesis-related ( PR ) gene PR1b1 [59] was examined with qRT-PCR . As shown in Figure 7K , whereas B . cinerea-induced expression levels of PR1b1 were reduced in spr8 plants than those in WT plants , expression levels of PR1b1 were much higher in OE-5 plants than those in WT plants , suggesting that the resistance of plants to pathogen is correlated with the expression levels of defense-related genes .
Here , we provide several lines of evidence demonstrating that the wound response defect of the tomato spr8 mutant results from a mutation in TomLoxD that is required for wound-induced JA biosynthesis . First , positional cloning studies reveal that spr8 plants harbor a dominant negative mutation in TomLoxD , a 13-lipoxygenase that catalyzes the oxygenation of the polyunsaturated fatty acid linolenic acid , which is the metabolic precursor of JA . Second , spr8 leaves accumulate very little JA in response to wounding . The deficiency in wound-induced JA biosynthesis accounts for the defective wound-induced PIs expression in spr8 plants and is consistent with the fact that the wound response phenotype of the mutant can be rescued by exogenous JA . These results lead us to conclude that TomLoxD is responsible for the majority of wound-induced JA production in tomato leaves . It is worth to note that the spr8 mutation affects a highly conserved Pro residue ( Pro598 ) in the lipoxygenase domain of TomLoxD ( Figure S4 ) . As an α-amino acid , Pro contains a distinct cyclic structure and therefore this amino acid exhibits an exceptional conformational rigidity compared to other amino acids [48] . In this context , it is reasonable to speculate that the spr8 mutation affects the formation of the secondary structure of the TomLoxD protein and hence impairs its activity . Indeed , our data support that , even though the expression of TomLoxDP598L is still responsive to mechanical wounding ( Figure 5A ) , this mutant version of TomLoxD impairs wound-induced JA biosynthesis ( Figure 5B ) . Considering that the spr8 mutation occurs in the C-terminal α-helices domain of TomLoxD ( Figure 4D ) , it is most likely that , in spr8 plants , the TomLoxDP598L protein still can bind the substrate ( i . e . , linoleic acid ) as the WT TomLoxD does , but this mutant protein loses its catalytic activity . Competition between TomLoxD and TomLoxDP598L for substrate binding could underlie that spr8 acts genetically as a semi-dominant mutant . As in other higher plants , in tomato lipoxygenases are encoded by a gene family consisting of 6 members ( Figure 4E ) . It has been shown that TomLoxA , TomLoxB , TomLoxC and TomLoxE are mainly expressed in fruits during development and ripening [60] . Among them , TomLoxC is specifically involved in the generation of C6 aldehydes and alcohols , which are important constituents of volatile flavor of tomato fruits [61]–[63] . The expression of TomLoxD and TomLoxF is stimulated by the non-pathogenic rhizobacteria Pseudomonas putida BTP1 and these two genes are likely to be involved in rhizobacteria mediated-induced systemic resistance [64] . The deduced amino acid sequence of TomLoxD show high similarity to several chloroplast-localized lipoxygenases in Arabidopsis that have been shown to be involved in JA biosynthesis . Among them , LOX3 and LOX4 are involved in male fertility [49] , [65] whereas LOX2 is specifically involved wound response [66] , [67] . TomLoxD also shows high sequence similarity to the maize TASSELSEED1 ( TS1 ) protein , which also encodes a plastid-localized lipoxygenase and plays a critical role in flower development and sex determination [68] . Here , we show that the tomato TomLoxD gene is specifically involved in the wound response , but shows minor , if any , effect on general plant growth and flower development ( Figure S1 ) . Taken together , these studies indicate that individual lipoxygenase isoforms are differentially regulated and have distinct physiological functions . Over two decades ago , Ryan and colleagues discovered the role of JAs in regulating defense gene expression in tomato [1] , [20] , [22] . Since then an ever growing body of evidence supports the view that the intracellular levels of JA plays a major role in controlling the strength of JA responses . Genetic engineering of plant cells for elevated endogenous JA levels therefore provides a strategy for increasing JA-dependent defenses . Indeed , the Ryan group showed that 35S::PS plants contain elevated JA levels and constantly express a spectrum of defense-related proteins that provide protection against insect attack [15] , [69] , [70] . Other examples of genetic alterations that cause constitutive JA accumulation include overexpression of a mitogen-activated protein kinase in tobacco [71] and mutation of the cellulose synthase CeSA3 in Arabidopsis [72] , [73] . It is noteworthy that even though genetic engineering of the tomato PS gene or the Arabidopsis CeSA3 gene leads to increased JA-dependent resistance against insects or pathogens , the resulting transgenic plants show growth retardation and other physiological defects in normal growth conditions [15] , [72] , [73] , which may limit the application potential of these genes in crop protection . Attempts to increase endogenous JA levels and thus JA-dependent resistance by overexpression of individual JA biosynthetic genes in tomato and other plants have met with limited success [18] , [33] , [74] , a plausible explanation is that the JA levels are mainly controlled by substrate availability [47] , [75] , [76] . In contrast to these unsuccessful examples , we show here that TomLoxD-OE plants exhibited increased expression levels of wound-induced defense-related genes and are more resistant to H . armigera . TomLoxD-OE plants also displayed enhanced resistance to the necrotrophic pathogen B . cinerea . These results indicated that genetic manipulation of TomLoxD leads to enhanced resistance of tomato plants to arthropod herbivores and microbial pathogens . It is important to note that in the absence of insect attack or pathogen infection , the overall growth and fertility of TomLoxD-OE plants were essentially comparable with those of WT plants ( Figure S1 ) , indicating there was no fitness cost associated with overexpressing TomLoxD in our growth conditions . This is important because the maintenance of constitutive proteins or the continuous mounting of defenses often has severe impacts on plant growth or fertility [77] . Because the overexpression of TomLoxD does not impose a significant fitness cost to the plant , the TomLoxD-OE plants are viable candidates for field trials to improve insect and pathogen resistance in crop tomato . Enhanced expression of defense-related genes in TomLoxD-OE plants is only observed after mechanical wounding , insect attack or pathogen infection suggests that the activation of the TomLoxD activity is regulated by the JA signaling . Indeed , we found that the wound-induced expression of TomLoxD is under the direct regulation of SlMYC2 , the functional homolog of the Arabidopsis MYC2 , a master regulator of JA-responsive gene expression . These findings are consistent with the long-standing observations that JA-signaling and synthesis form an apparent positive feedback regulatory loop [25] , [26] , [78] . It is also possible that the activity of TomLoxD for wound-induced JA biosynthesis is under posttranscriptional modification and that this modification is regulated by environmental stimuli including wounding , insect attack or pathogen infection . Alternatively , these environmental stimuli could lead to the accumulation of more substrates available for TomLoxD . Given the application potential of TomLoxD for crop protection , it is of significant in future studies to further explore the functional mechanisms of TomLoxD in wound-induced JA biosynthesis .
Tomato ( Solanum lycopersicum L . ) cv Castlemart ( CM ) was used as the wild-type ( WT ) for all experiments . The plant material 35S::PS used in this study was previously described [15] , [17] , [29] . Tomato seedlings were grown in growth chambers and maintained under 16 h of light ( 200 µE m−2 s−1 ) at 28°C and 8 h of dark at 18°C and 60% relative humidity . Mutagenesis of 35S::PS plants with ethyl methanesulfonate ( EMS ) and the isolation of suppressor of prosystemin-mediated responses ( spr ) mutants were performed as previously described [17] , [29] . spr8 is one of the identified mutant lines and is deficient in both PPO activity and PI-II protein accumulation . The original spr8 mutant in the 35S::PS genetic background was backcrossed to tomato cv CM as previously described [18] . The identified homozygous spr8/spr8 mutant plants were crossed to the WT and F1 plants were allowed to self-pollinate . The wound response phenotype of F1 and F2 plants was assessed by measuring PI-II accumulation following wounding treatment . Map-based cloning procedures similar to those described [18] , [79] were used to identify the Spr8 locus . A homozygous spr8 plant ( S . lycopersicum ) was crossed to the wild tomato species S . pennellii ( LA716 ) , and the resulting F1 plant was backcrossed to the spr8 parental line to generate a BC1 mapping population . The wound-response phenotype of individual BC1 plants was scored by measuring PI-II protein levels in response to mechanical wounding , as described above . Using the BC1 population described above , bulked segregant analysis was used in combination with simple sequence repeats ( SSR ) analysis to identify molecular markers linked to Spr8 . Equal amounts of genomic DNA from10 randomly selected wound-responsive ( i . e . , wild-type ) and 10 nonresponsive ( i . e . , mutant ) BC1 plants were pooled to construct a wild-type DNA bulk ( B+ ) and a mutant DNA bulk ( B− ) , respectively . Rough mapping using the 20 BC1 plants indicated that the target gene is linked to the marker TES0023 on the long arm of chromosome 3 . Analysis of linkage between Spr8 and known SSR markers in this region demonstrated that Spr8 is located between TES0023 and TES1203 . A high-resolution genetic map of the Spr8 region was constructed by scoring 354 BC1 plants for recombination events within the SSR601-Spr8-M140 interval in the scaffold SL2 . 40sc03701 of the sequenced tomato genome . Sequence analyses of genes in this interval revealed a C-to-T mutation in the TomLoxD gene . DNA primers for molecular markers used in map-based cloning were listed in Table S1 . For complementation analysis , the 35Spro:TomLoxD-GFP construct was introduced into the spr8 plants using Agrobacterium tumefaciens-mediated transformation for the complementation analysis . The TomLoxD-RNAi and 35Spro:TomLoxDP598L-GFP constructs were introduced into WT plants using Agrobacterium tumefaciens-mediated transformation . DNA constructs for plant transformation were generated following standard molecular biology protocols and Gateway ( Invitrogen ) technology . Full-length coding sequence of TomLoxD was amplified with Gateway-compatible primers . The PCR product was cloned by pENTR Directional TOPO cloning kits ( Invitrogen ) and then recombined with the binary vector pGWB5 ( 35S promoter , C-GFP ) to generate the 35Spro:TomLoxD-GFP construct . Similarly , we generated 35Spro:TomLoxDP598L-GFP construct , which was amplified from spr8 cDNAs , using the same primers as 35Spro:TomLoxD-GFP construct . Full-length coding sequence of SlMYC2 was also cloned into the pGWB17 vector ( 35S promoter , C-4myc ) to generate the 35Spro:SlMYC2-4myc constructs . To generate a TomLoxD-RNAi construct , fragments of the TomLoxD open read frame ( 106–570 bp ) , which were amplified from the cDNAs , were digested by XhoI and SpeI , and then inserted into XhoI-SpeI sites and SalI-XbaI sites in PUCCRNAi vector by steps . So this second ligation inserts the PCR product was in inverted orientation with respect to first cloned fragment , yielding an inverted repeat separated by the first intron fragment of GA20 oxidase from potato . The two reversed repeated DNAs were transferred to pCAMBIA-1301 ( CAMBIA ) from PUCCRNAi by PstI digestion . The plasmid pCAMBIA-1301 had been modified by adding a CaMV 35S promoter . Similarly , the SlMYC2-RNAi construct was performed . All primers used for DNA construct generation are listed in Table S3 online . The above constructs were then transformed into Agrobacterium tumefaciens strain AGLO and used to transform tomato cotyledon explants as described previously [18] . Transformants were selected based on their resistance to hygromycin . Homozygous T3 or T4 transgenic seedlings were used for phenotype and molecular characterization . The wound response of tomato plants was determined using a radial immunodiffusion assay for the detection of PI-II accumulation in leaf tissue as previously described [11] , [36] . Two-leaf-stage ( 16-day-old ) seedlings were used for the wounding treatment as described [29] and then the wounded leaf ( local response ) and the unwounded leaf ( systemic response ) were harvested separately to assay PI-II protein level . For wounding treatment , 16-day-old seedlings were wounded with a hemostat across the midrib of all leaflets on the lower leaf and the upper leaf . Then , the same leaflets were wounded again , proximal to the petiole . Wounded plants were incubated under continuous illumination conditions . For each time point of sampling , five whole plants leaves were harvested for the extraction of RNAs . Systemin feeding experiments were performed using 16-day-old tomato seedlings as previously described with minor modifications [18] , [28] , [29] . Briefly , 2 . 5 pmol systemin was diluted from stock solutions into 300 µL 15 mM sodium phosphate , pH 6 . 5 , prior to use . Tomato seedlings were excised at the base of the stem and placed in 0 . 5 mL microfuge tubes containing 300 µL dilutions . When >90% of the elicitor solution had been imbibed ( approximately 2 hours ) , plants were transferred to glass vials containing 20 mL of water , and incubated in a Lucite Box under continuous light . Twelve hours later , leaf tissues of five plants were pooled for RNA extraction and gene expression assays . Control plants were fed with sodium phosphate buffer . Systemin was commercially synthesized by Shanghai Sangon Biological Engineering & Technology and Service Co . Ltd ( Shanghai , PR China ) . Sixteen-day-old tomato seedlings were treated with MeJA as described previously [80] . Control plants were incubated in a separate container in which ethanol was applied to cotton wicks . Twelve hours later , leaf tissues of five plants were pooled for RNA extraction . MeJA was purchased from Sigma-Aldrich . For qRT-PCR analysis , leaf tissues were harvested and frozen in liquid nitrogen for RNA extraction . RNA extraction and qRT-PCR analysis were performed as previously described [50] . Expression levels of target genes were normalized to those of the tomato Actin2 gene . Primers used to quantify gene expression levels are listed in Table S2 . To examine the general pattern of trichome distribution on the adaxial surface of leaves , small pieces of tissue ( 5×5 mm ) , on the same base region of the third leaves from bottom to upper , were fixed , dehydrated , critical point dried in CO2 , and coated with a film of gold as described [81] . Observations were performed with a HITACHI S-3000N scanning electron microscope ( Japan ) at an accelerating voltage of 15 kV . The density of type VI trichomes on the adaxial surface of leaves was determined by counting trichomes with a dissecting microscope equipped with a stage micrometer . All measurements were performed on WT and spr8 plants grown side by side under the same growth conditions . Five-week-old plants were used to isolated type VI trichomes of leaves to obtain trichome exudates as previously described with minor modified [43] . Briefly , 1 , 000 heads of Type VI glandular trichomes were selectively collected with a stretched-glass pipette and dissolved into 200 µL methyl tert-butyl ether ( MTBE , Sigma ) to analysis the chemical structures of compounds by GC-MS as described [43] . Different concentrations of external standards were run under the same GC conditions to develop standard curves to quantify volatiles ( 2-carene for monoterpenes , β-caryophyllene for sesquiterpenes ) . General procedures for rearing and handling cotton bollworm ( Helicoverpa armigera ) were described previously [18] , [79] . The average larval weight at the beginning of the feeding trial was ∼5 mg . After termination of the feeding trial , PI-II protein accumulation in the remaining leaf tissues was measured [11] , [36] , as was the weight gain of larvae reared on both of the host genotypes . Detached leaves of five-week-old plants were inoculated as previously described [51] . For qRT-PCR experiments , the inoculation tests were performed in planta as described [58] . The same experiment was done with mock-pretreated plants as control . After inoculated for different times , the samples were then harvested for RNA extraction . The BLAST search program [82] was used for sequence analysis . The software ClusterX and T-coffee ( http://www . ebi . ac . uk/Tools/t-coffee/ ) were used for sequence alignment . The phylogenetic relationship of TomLoxD in plants is inferred from protein sequences using a Bayesian approach in MrBayes [83] . The node labels are measures of support , which indicate the proportion of trees in the posterior distribution to containing the node . For JA content measurement , 16- to 18-day-old plant leaves were wounded as described above . Approximately 200 mg leaf tissue ( fresh weight ) from five different plants was pooled for JA quantification as described previously [84] . Leaf tissues were also harvested from unwounded plants as controls . ChIP assays were performed following a published protocol [50] , [51] , [54] , [85] with minor modifications . Briefly , 1 hour after wounding treatment , 2 . 0 gram of 16-day-old 35Spro:SlMYC2-4myc plant leaves were harvested and cross-linked in 1% formaldehyde for ChIP experiment . myc antibody ( Millipore ) was used to immunoprecipitate the protein-DNA complex , and the precipitated DNA was purified using a PCR purification kit ( Qiagen ) for PCR analysis . Chromatin precipitated without antibody was used as negative control , while the isolated chromatin before precipitation was used as input control . Primers used for ChIP-PCR are listed in Table S4 online . For plasmid construction of maltose binding protein ( MBP ) fusions with SlMYC2 , the cDNA was amplified and cloned into the pMAL-c2 vector ( New England Biolabs , Beverly , MA ) via BamHI and PstI restriction sites . The MBP-SlMYC2 recombinant protein was expressed in the BL21 Escheichia coli ( E . coli ) strain and purified by binding onto an amylose resin ( New England Biolabs ) column , according to the instructions provided by the manufacturer . The 50-bp TomLoxD promoter probes containing G-box-like motif at the -369 site were synthesized and labeled with biotin at the 3′ end ( Invitrogen ) , which containing the same sequences as that of the competitor probes without biotin-labled , while the mutated labeled probes were deleted the G-box-like motif . EMSA assays were performed using a LightShift Chemiluminescent EMSA kit ( Thermo Scientific ) as described [54] . Probe sequences are shown in Table S4 online . The transient expression assays were performed in N . benthamiana leaves as previously described [51] , [54] . The TomLoxD promoter was amplified and cloned into the pCAMBIA1381-Z ( CAMBIA ) via EcoRI and PstI restriction sites to generate the reporter construct TomLoxDpro:LUC . The SlMYC2 effector construct was the above-described 35Spro:SlMYC2-4myc . We used a low-light cooled CCD imaging apparatus ( NightOWL II LB983 with indigo software ) to capture the LUC image and to count luminescence intensity . The leaves were sprayed with 100 mM luciferin and were placed in darkness for 3 min before luminescence detection . For plasmid construction of 35Spro:TomLoxD-GFP , the full length cDNA was amplified and cloned into the pGFP-2 vector [86] via XhoI and KpnI restriction sites to generate 35Spro:TomLoxD-GFP . Protoplast isolation and analysis of the subcellular location of transiently expressed GFP fusions by confocal fluorescence microscopy were performed as described [87] . Alexander's triple staining was employed to measure viability of pollens , which were freshly harvested , as described previously [88] . The accession number of the sequenced tomato genome for the scaffold containing the Spr8/TomLoxD gene is SL2 . 40sc03701 . The accession number from SGN database as following: TomLoxD ( Solyc03g122340 ) ; SlMYC2 ( Solyc08g076930 ) . Sequence data from this article can be found in the in the Arabidopsis Genome or GenBank databases under accession number as following: Arabidopsis thaliana AtLOX1 ( AT1G55020 ) , AtLOX2 ( AT3G45140 ) , AtLOX3 ( AT1G17420 ) , AtLOX4 ( AT1G72520 ) , AtLOX5 ( AT3G22400 ) , AtLOX6 ( AT1G67560 ) ; Solanum lycopersicum TomLoxA ( P38415 ) , TomLoxB ( P38416 ) , TomLoxC ( AAB65766 ) , TomLoxD ( AAB65767 ) , TomLoxE ( AAG21691 ) , TomLoxF ( NP_001234259 ) ; Zea mays ZmTS1 ( ACL81190 ) ; Solanum tuberosum LOXH3 ( CAA65269 ) , StLOXH1 ( CAA65268 ) , STLOX ( AAD09202 ) , POTLX-3 ( AAB67865 ) , St13s-LOX2-1 ( O24370 ) , St13s-LOX3-1 ( O24371 ) ; Nicotiana tabacum NtLOX ( CAA58859 ) ; Glycine max Gm13-LOX3-1 ( XP_003528556 ) ; Oryza sativa Japonica Group OsLOX6 ( NP_001049158 ) ; Rattus norvegicus RnLOX3 ( NP_001099263 ) ; Mus musculus Mm5-LOX ( NP_033792 ) ; Homo sapiens HsLOX3 ( CAC12843 ) .
|
Plants have evolved sophisticated strategies to defend themselves against insect attack . Wound-inducible proteinase inhibitors ( PIs ) in tomato ( Solanum lycopersicum ) provide an attractive model to understand the signal transduction events leading from localized injury to the systemic expression of defense-related genes . A wealth of evidence indicates that the peptide signal systemin and the phytohormone jasmonic acid ( JA ) work together in the same signaling pathway to activate the expression of PIs and other defense-related genes . We have been using a genetic approach to dissect the systemin/JA signaling pathway and to discover important genes that can be used for crop protection . Here we report the characterization of the suppressor of prosystemin-mediated responses8 ( spr8 ) mutant , which is defective in wound-induced defense gene expression and therefore is more susceptible to insect attack . We demonstrate that spr8 defines the TomLoxD gene , which encodes a chloroplast-localized lipoxygenase involved in wound-induced JA biosynthesis . Further , we demonstrate that genetic manipulation of Spr8/TomLoxD leads to increased plant resistance against insect attack and pathogen infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Role of Tomato Lipoxygenase D in Wound-Induced Jasmonate Biosynthesis and Plant Immunity to Insect Herbivores
|
Resistance to pyrethroids and to the organophosphate temephos is widespread in Brazilian populations of the dengue vector , Aedes aegypti . Thereof , since 2009 Insect Growth Regulators are employed as larvicides , and malathion is used against adults . We performed laboratory selection with malathion of two A . aegypti field populations initially susceptible to this organophosphate but resistant to temephos and deltamethrin . A fixed malathion dose inducing at least 80% mortality in the first generation , was used throughout the selection process , interrupted after five generations , when the threshold of 20% mortality was reached . For each population , three experimental and two control groups , not exposed to insecticides , were kept independently . For both populations , quantitative bioassays revealed , in the selected groups , acquisition of resistance to malathion and negative impact of malathion selection on deltamethrin and temephos resistance levels . In the control groups resistance to all evaluated insecticides decreased except , unexpectedly , to deltamethrin . Analysis of the main resistance mechanisms employed routine methodologies: biochemical and molecular assays for , respectively , metabolic resistance and quantification of the NaV pyrethroid target main kdr mutations at positions 1016 and 1534 . No diagnostic alteration could be specifically correlated with malathion selection , neither with the unusual deltamethrin increase in resistance levels observed in the control groups . Our results confirm the multifactorial character of insecticide resistance and point to the need of high throughput methodologies and to the study of additional field vector populations in order to unravel resistance mechanisms .
Aedes aegypti ( Linnaeus , 1762 ) is the main vector of dengue , Zika and chikungunya viruses , which are currently serious public health problems in Brazil [1] . Specific medications are not available for any of these viruses . One dengue vaccine was recently approved but it confers only partial protection [2]; other vaccines against dengue and Zika viruses are yet in the clinical trials phase [3–5] , and efforts in this regard for the chikungunya virus are just beginning [6] . Moreover , since 2016 , Brazil faces a public health emergency with respect of yellow fever [7] . Although its transmission is silvatic , by Haemagogus and Sabethes mosquitoes [8] , with the low vaccination coverage of the population there is an imminent risk of reurbanization of the virus , and A . aegypti is its main urban vector [9 , 10] . Except for the yellow fever virus , in the current scenario prophylactic measures depend on control initiatives aimed at decreasing vector density [11 , 12] . Although the most recommended and effective measures are the mechanical control and society awareness , chemical insecticides are still widely used tools . However , the intense exposure of vector and pest populations to insecticides can select characteristics and mechanisms that confer resistance to these xenobiotics [13] . Vector populations resistant to insecticides are a threat to the success of control programs that prioritize chemical control [14 , 15] . Thus , monitoring the susceptibility status of natural populations to the main employed insecticides is one of the pillars of vector control programs . This is a way of guiding control actions and preventing resistance of reaching a threshold considered of risk [12 , 16] . The organophosphate ( OP ) temephos has been used in the control of A . aegypti larvae in Brazil since 1967 . Resistance to temephos , monitored in vector populations throughout the country , started in 2000–2001 and increases since then . In 2009 , due to the spread of resistance [17] , temephos ceased to be the first-choice larvicide , being replaced by chitin synthesis inhibitors ( CSI ) [15 , 18] . Similar to the control of larvae , OP were also used against adult mosquitoes up to 2001 , when pyrethroids ( PY ) replaced them [19] . However , shortly thereafter , reduction of PY susceptibility of various field populations started to be identified [20 , 21] , a process that resulted in the decision of interrupting the use of this class of insecticides in the control of A . aegypti adults in 2009 [18 , 22] . The Brazilian Ministry of Health ( MoH ) adopts only insecticides recommended by WHO [23] for the control of A . aegypti . Taking into account resistance of adults to PY insecticides , the only viable alternative for ultra low volume applications was malathion [23] . Although malathion is also an OP , such as temephos , the structure of both compounds is distinct: while temephos is a closed-chain phosphorothionate , malathion , an open-chain compound , is a phosphorodithioate , and bears two sulfur atoms attached to the central phosphorus , rather than just one [24] . Regarding the perspective of vector control , such differences are important , since they contribute to elicit different resistance mechanisms , as is the case with Anopheles mosquitoes , for example [25 , 26] . Indeed , in Brazil , in contrast to the widespread A . aegypti resistance to temephos , registers of altered malathion susceptibility were scarce , and only derived from qualitative bioassays [27 , 28] . When PY were replaced by malathion in the control of adults , quantitative assays confirmed the susceptibility of Brazilian natural A . aegypti populations to this OP ( Braga TA , personal communication ) , corroborating the pertinence of its use in the field . Taking this scenario into account , the anticipation of a potential malathion resistance event , as well as the elucidation of the possible mechanisms involved , would be of great value for the definition of rational strategies that could prolong this OP as a viable alternative for the control of A . aegypti adult specimens in Brazil . This study describes the selection with malathion , in the laboratory , of two A . aegypti field populations resistant to temephos and to the PY deltamethrin . Alterations in the malathion susceptibility profile were investigated , as well as the consequences of malathion selection on the susceptibility profile of the resulting samples to the main insecticides used in the country against the vector: temephos , the CSI diflubenzuron and deltamethrin . Collection of A . aegypti population from the municipality of Aracaju , at Sergipe State ( SE ) was made in 2012 when the MoH applied both temephos and CSI against larvae , and only PY to control adults . Crato mosquitoes , from Ceará State ( CE ) , the higher resistant population , was collected in 2013 when temephos applications had just been replaced by CSI , and control of adults employed both PY and malathion . More details regarding the history of insecticides use in both localities are shown elsewhere [29] . Our major aim was to investigate if selection with malathion for a limited number of generations would be enough to alter the susceptibility profile of the exposed populations . In addition , it would be of interest to evaluate to what extent different vector populations , bearing distinct backgrounds regarding challenges with insecticides and other xenobiotics , would be impacted by malathion selection .
Sampling of A . aegypti eggs representative of the evaluated municipalities was done with ovitraps [27 , 30] . In the laboratory , mosquito colonies were started with 349 positive ovitraps from Aracaju / SE installed in 2012 , and 272 positive ones from Crato / CE , installed in 2013 [29] . Egg hatching and larval rearing were carried out as routinely [27] . The resulting A . aegypti adults were then identified and used to obtain F1 eggs . Although the initial objective was to employ F1 specimens to characterize the resistance and the resistance mechanisms of the original populations [29] and also to proceed with the selection , the amount of eggs obtained was not enough . Therefore , F2 generation specimens were used in the selection procedure . In addition to malathion selection , the susceptibility status to the main insecticides applied by the Brazilian Dengue Control Program was analyzed . The study adopted technical grade compounds: the OP temephos ( Pestanal; Sigma-Aldrich Brasil Ltda , São Paulo , SP , Brazil ) and malathion ( Cheminova Brasil Ltda , São Paulo , SP , Brazil ) , the PY deltamethrin ( Pestanal; Sigma-Aldrich Brasil Ltda ) and the CSI diflubenzuron ( Pestanal; Sigma-Aldrich Brasil Ltda ) . A malathion dose capable of killing about 80% of specimens in the first generation was used . This same dose was kept fixed until the mortality was only 20% . Three biological replicates of the selection ( S1 , S2 and S3 ) were done per population , without exchange of specimens between the replicates . For each population two control groups ( C1 and C2 ) were also included , kept simultaneously in the same conditions , but without any contact with insecticides . Malathion concentrations were defined in previous trials: 0 . 06229 mg/L for Aracaju and 0 . 09876 mg/L for Crato . For the first generation of selection , at least 5 , 000 larvae of each population were used per replicate , and 3 , 000 specimens were used to initiate subsequent generations of each independent experimental group . Synchronized larvae were used in the whole selection process: hatching of A . aegypti eggs was induced for at most one hour in a BOD incubator at 28°C . Larvae were then kept at 26 ± 2°C , at the density of 500 larvae per basin , in 1 L dechlorinated water and 1 g of grounded cat food ( Friskies , Purina/Camaquã/RS ) was provided . Under these conditions , larvae collected 72 hours after hatching were essentially L3 ones , and approximately 5% of the specimens had moulted to L4 . Basically , for each strain and each replica , several groups of 100 larvae , 72 hours old , were placed in plastic cylindrical basins ( 10 X 12 cm ) with 250 ml of dechlorinated water and malathion ethanolic solution ( for Aracaju and Crato samples , respectively 346 and 548 . 6 μl of a 45 mg/l malathion solution were employed ) or equivalent volume of ethanol , in the case of control replicas . Larvae were fed with 0 . 2 g of grounded cat food added each three days . Exposure to malathion was continuous up to pupation . Pupae were daily transferred to cylindrical cardboard cages ( 17 X 18 cm ) . Upon eclosion , adults were fed ad libitum ( except before the blood meal ) with a 10% sugar solution , replaced three times a week and kept in a temperature and humidity-controlled insectary ( 26 ± 1°C; 80 ± 10% rh ) . To obtain eggs , females were weekly deprived of the sugar solution for 18–24 hours , and anaesthetized guinea pigs were then offered as blood source according to the “Formulary for laboratory animals” [31] . After five generations , all replicates exposed to malathion exhibited less than 20% mortality in the presence of the insecticide , and the experiment was interrupted . Quantitative bioassays of larvae with temephos followed the parameters and procedures described by WHO [32] . Bioassays with diflubenzuron were made according to Martins et al . [33] . Adult bioassays with PY and malathion were done according to an adaptation of the insecticide impregnated papers methodology [34 , 35] . In all cases , three to four assays were performed for each population on different days . As an internal control , simultaneous trials were also carried out with the Rockefeller strain , a reference of insecticide susceptibility [36] . In each bioassay A . aegypti specimens were exposed to an insecticide spectrum of at least six concentrations . Larvae were exposed to temephos for 24 hours and to diflubenzuron until death or emergence of the last adult specimen of the control group , as previously defined by Braga et al . [37] . Adults were exposed to either deltamethrin or malathion for 1 hour and then transferred to a recovery chamber , with no insecticide , where they remained for another 24 hours , when mortality was recorded . For each assay with larvae , 320 to 640 specimens were used , totaling 960–1 , 920 specimens per insecticide tested by biological replicate ( S1-S3 , C1 , C2 ) . In the case of adults , 360–480 specimens were used per assay , corresponding to 1 , 080–1 , 440 specimens for each insecticide per biological replicate . The TaqMan method was used to identify kdr mutations in the NaV gene in the post-selection samples , as previously described for pre-selection populations [29] . For each replica ( C1 , C2 , S1 , S2 , S3 ) , 30 individual males were used . Analysis of Aracaju samples was done with F7 specimens ( offspring of the last generation of selection in the replicates S1-S3 ) . In the case of Crato , analysis of C1 and C2 samples was done with F6 , and S1-S3 replicates were investigated with F7 mosquitoes . Independent reactions were done for the substitutions Val1016Ile and Phe1534Cys . For each position , 1 μl , equivalent to 0 . 5% of the DNA content extracted from each specimen , was used in a 10 μl final reaction volume . One-day-old adult females were evaluated according to methodology adapted from WHO and the US Centers for Disease Control and Prevention ( CDC ) [38–40] . The same protocol , with some additional adaptations , was applied to larvae [41] . Activity of mixed function oxidases ( MFO ) , esterases ( EST ) and glutathione S-transferases ( GST ) were quantified . Three substrates were used for ESTs: α-naphthyl , β- naphthyl and ρ-nitrophenyl acetates , related respectively to α-EST , β-EST and ρNPA-EST activities . For Acetylcholinesterase ( Ace ) , the OP target site , both total activity ( AChE ) and the remaining activity after inhibition by propoxur ( AChI ) were evaluated . In order to calculate the specific enzymatic activities , total protein content of each sample was quantified using the Bio-Rad protein reagent ( catalog number: 500–0006 ) . At least three assays were accomplished on different days for all enzymes . In each assay 40 individual specimens of each experimental replica were evaluated . Simultaneously , as an internal control of the assays , five specimens of the Rockefeller strain were also analyzed . In the case of bioassays , adult emergence inhibition ( EI ) or lethal concentrations ( LC ) were calculated , respectively for diflubenzuron and neurotoxic insecticides , by probit analysis using the Polo-PC software ( LeOra Software , Berkeley , CA ) [42] . Results of the quantitative assays were expressed as the ratio of resistance between the LC ( or EI ) of the experimental group under test and the Rockefeller equivalent measure ( RR ) . Another index , 'selection ratio' ( SR ) , was also calculated , by comparison of LC ( or EI ) values of C and S groups with their corresponding parental ( P ) population ones [29] . Additionally , we assessed the 95% confidence intervals overlap range between the post- and pre-selection LC of each experimental sample . The criterion used in Brazil to classify the resistance status of A . aegypti populations to temephos [43] was also applied to the other insecticides evaluated in this study . According to this criterion , populations with RR95 higher than 3 . 0 are classified as resistant , and there is a recommendation for replacement of the insecticide compound used in the field . Individual NaV genotypes , as well as the allelic and genotypic frequencies of each population , were calculated based on the variations in positions 1016 and 1534 , both on the same gene , as described elsewhere [44] . The 95% confidence intervals overlap range between the post- and pre-selection kdr allelic frequencies was also evaluated . Enzymatic activities were classified according to a previously established criterion , based on the use of the 99th percentile of a reference strain as the cutoff point . In this case , references corresponded to the parental strains , from Aracaju and Crato [29] . After calculation of their 99th percentiles for each enzyme class activity , rates of corresponding specimens in the C and S groups above this value were estimated . The activities were classified as unaltered , altered or highly altered in cases where these rates were respectively less than 15 , between 15 and 50 , or above 50% [43 , 38] . In particular , in the case of the Ace inhibition assay with propoxur ( AChI ) , the WHO criterion establishes that remaining activity exceeding 30% is indicative of resistance to OP insecticides [40] . All enzyme activity profiles were also compared using the Kruskal Wallis nonparametric test and the Dunn's multiple comparison post test ( using the graphpad prism version 5 . 0 ) . The blood feeding was done according to the Brazilian guidelines described in “The Brazilian legal framework on the scientific use of animals” [45] , supported by a protocol approved by the "Ethics Committee in the Study of Animals" ( CEUA/Fiocruz 2008 ) , licenses L-011/09 and LW-20/14 .
In the first generation of selection with a fixed malathion dose , respectively 20 . 4 and 11 . 8% of Aracaju and Crato individuals survived . In the course of the selection process , the Aracaju survival percentage increased more evenly than that of Crato . Nevertheless , both populations reached the threshold of 80% survival ( dotted line in Fig 1 ) , defined as the time of selection interruption , in the 5th generation ( F6 ) . For the population of Crato , exposure to malathion was repeated with one additional generation , when 90 . 9% of survival was detected . After that , the potential impact of laboratory selection on the resistance status of malathion itself ( in larvae and adults ) and on other compounds used in the A . aegypti control in the country was assessed . It is ought to mention that mortality in the control groups did not exceed 1% during the whole process . Table 1 summarizes the impacts of malathion selection and of adaptation to laboratory breeding on the RR profile to several insecticides employed in A . aegypti control . In order to facilitate a direct comparison , results of the original evaluations with the pre-selected populations [29] were included . Bold values in Table 1 point to resistance when the Brazilian MoH classification criteria , previously defined for temephos , is used ( RR95 > 3 . 0 , see Methods ) . Detailed results are presented in S1–S5 Tables , including comparison of mortality values obtained for C and S groups with their corresponding parental ( P ) populations ( the 'selection ratio' , or SR analysis ) . Both analytical approaches pointed out the effectiveness of the laboratory selection promoted in the study . Although SR data are slighter than RR ones: ( a ) both malathion RR and SR increased in the larvae of virtually all selected groups—malathion SR values increased in the range of 50% and more than twice for , respectively , Aracaju and Crato selected larvae; ( b ) in general , there was no overlap of CL values between C and S groups , pointing to a sound effect of malathion selection ( S1 Fig ) ; ( c ) for Crato , the selective pressure reverberated in the adult stage to the point that it could also be classified as resistant after selection , when the 'MoH RR95 criterion' was used . Malathion SR values increased in the range of 50% for Crato adults . Groups maintained without insecticide showed a slight reduction of malathion RR in the larval stage ( which were already low ) , and a discrete increase in the adult stage , while SR revealed a slight increase in both stages ( S1 and S2 Tables ) . In general , for each population , C replicas , as well as S ones , had an equivalent performance , although reared independently . Laboratory rearing in the absence of insecticides ( groups C1 , C2 ) was enough to slightly reduce the initial temephos resistance levels exhibited by parental populations . Selection with malathion did not induce cross-resistance with temephos , although both are OP ( Tables 1 and S3 ) . In contrast , temephos resistance decrease was more noticeable in malathion selected groups than in the control ones–both RR and SR indexes suggested such a trend . In particular , temephos resistance reduction was more pronounced in Crato , the population with the highest initial RR . When the present study started , diflubenzuron had been recently introduced in the country to control A . aegypti larvae . Resistance indexes for this CSI compares adult emergency inhibition rates ( EI ) , and not lethal concentrations . When compared to the original populations , although some differences in RR and SR for DFB were observed between C and S groups , all values obtained were considered compatible with a susceptible status ( Tables 1 and S4 ) . All samples were considered highly resistant to deltamethrin . Surprisingly , the simple laboratory rearing , in the absence of any insecticide ( groups C1 , C2 ) , exacerbated PY resistance levels in both populations ( Tables 1 and S5 ) . In particular , the SR index confirmed the marked deltamethrin resistance increase in the control groups , mainly in Aracaju ( S5 Table ) . This increase in deltamethrin resistance , however , was not observed in the S1-3 groups , selected with malathion , independently of the analytical method adopted ( RR , SR and 95% confidence limits overlapping ) . In opposition , malathion selected groups , and particularly Crato mosquitoes , exhibited reduction of deltamethrin resistance levels . In general , when compared to the parental samples and as judged by slope values , homogeneity of all evaluated groups remained unaltered or increased ( S1–S5 Tables ) . Only with deltamethrin bioassays there was a trend towards increased heterogeneity ( except for Crato S groups ) . In this regard , the unusual and specific increase in deltamethrin resistance , noted in the control groups , should be considered . Table 2 shows the kdr genotypes at positions 1016 and 1534 of the NaV gene , the target site of PY insecticides . The 'wild type' susceptible allele , '1016 Val+ + 1534 Phe+' , was called 'S' . The allele mutated only at position 1534 ( 1016 Val+ + 1534 Cyskdr ) was named 'R1' and the other , with mutations at both positions ( 1016 Ilekdr + 1534 Cyskdr ) , 'R2' . Mutation exclusively at position 1016 was not found . The three alleles were detected in Aracaju and Crato . Since kdr mutations at positions 1016 and 1534 are recessive , PY resistance mediated by such changes is only expressed in homozygous individuals: R1R1 , R1R2 and R2R2 [44] ( highlighted in bold and summed in column ‘resistant genotypes’ in Table 2 ) . In Crato the wild-type ( S ) allele remained as the most frequent , followed by R1 , a pattern kept in all sample groups from that locality . On the other hand , R2 was the most frequent allele in Aracaju , followed by the S allele but this arrangement tended to reverse after laboratory rearing . Regarding the original populations , in both cases , an increase in the frequency of the S allele was observed in almost all samples , selected with malathion or simply reared in the laboratory , without insecticides ( note the IC 95% overlap range in S2 Fig ) . The initial kdr genotype frequencies , exhibiting PY resistance , were 43% for Aracaju and 32% for Crato ( 'resistant genotypes' column in Table 2 ) . In almost all cases , in both groups C and S , such frequencies decreased during the selection or rearing processes . Tables 3 , 4 and S7 show the activity profiles of the major classes of detoxification enzymes in larvae and adults of Aracaju and Crato mosquito populations . Acetylcholinesterase activity , target site of OP , was also quantified; in this case the total activity ( AChE ) and its profile of inhibition by propoxur ( AChI ) are presented . Tables 3 and 4 depict , respectively , the percentage of Aracaju and Crato individuals with activity above the 99th percentile of each corresponding Parental strain; S7 Table compares median values of the same samples , using the Kruskal-Wallis nonparametric test . For both populations more changes and higher activities were noted in adult mosquitoes , compared to the larval stage , in all C and S groups . The exceptions were α-EST and β-EST activities , which are higher in larvae ( S7 Table ) . In general , no differences were observed between control groups and those exposed to malathion which could be classified as diagnostic of the selected resistance . In relation to the parental samples , Aracaju adult females tended to present higher MFO and GST activities , the later being revealed by both criteria in control samples ( Tables 3 and S7 ) but detected only after comparison of median values in S ones ( S7 Table ) . In Crato , marked alterations of all classes of enzymes were also observed in both C and S females , when compared to the Parental population; however , GST alterations were more attenuated in relation to the other enzyme classes , being detected only by medians comparison ( Tables 4 and S7 ) . In the larval stage , no metabolic alterations were detected in Aracaju C and S samples , compared to the original population . In Crato larvae , α-EST EST activity was consistently enhanced and MFO alterations were noted in some samples when median values were compared , particularly in the malathion selected ones . Total activity of the OP target site , AChE , remained equivalent to the original samples . We also investigated Ace using the WHO susceptibility criterion: inhibition of 70% or more of Ace's activity with propoxur ( AChI ) ; in all cases the results are indicative of a susceptible enzyme ( Tables 3 , 4 and S7 ) . According to this criterion , Ace of both parental Aracaju and Crato populations has a susceptible profile .
This study deals with the response to laboratory rearing and to selection with malathion of two A . aegypti field populations . Their resistance status to the main insecticides employed by the Brazilian MoH as well as the potentially associated resistance mechanisms were evaluated . The outcome of malathion selection was considered consistent , since resistance to this OP increased in the three biological replicates of both Aracaju and Crato . This increase was seen in all the adopted analytical approaches—comparison with the Rockefeller susceptible strain or with the parental generation . Although comparison with the parental population ( SR values ) showed more slight increases than comparison with the susceptible reference strain ( RR ) , both selection efficacy and the functional significance of this result are confirmed since there is no overlap of CL ranges between the selected groups and the control samples , nor with the parental ones . Considering also the initial and final mortalities of the malathion selection experiment , which ranged from 80 to less than 20% in the six independently selected groups ( Fig 1 ) , we can infer that malathion selection appears to have had a relevant biological impact . The magnitude of such increases in resistance levels was also compatible with other records in the literature that subjected A . aegypti populations to different insecticides for a small number of generations [46 , 47] . Although selection has increased malathion resistance levels , resistance to temephos tended to decrease in both populations . Such reduction was more prominent in the samples selected with malathion than in the control groups . This situation was , to some extent , expected , considering the molecular structures of both insecticides and other reports in the literature [25 , 48–50] . Collectively , such data reinforce the idea that these two organophosphates elicit different mechanisms of resistance [25 , 50 , 51] . This also occurs , for example , with species of the genera Anopheles and Culex: resistance to temephos has been found to be derived from the copy number amplification of esterases genes , culminating in a greater number of available enzyme molecules [51 , 52]; on the other hand , malathion metabolic resistance is related to alterations in the coding region of a particular esterase , called 'malaoxonase' , resulting in a more efficient enzyme [26 , 53] . In fact , the two OPs differ in chemical structure , as stated in the Introduction . Such differences appear to result in the extremely slow hydrolysis of temephos , via esterases , a process that may last for days . In this case , it is even considered that temephos is sequestered [48 , 50] . As a consequence , esterase-mediated resistance to OP compounds , in general , would occur due to an increase in the amount of enzyme molecules , as is the case with Culex mosquitoes [51 , 54] . In our study , alterations in some specific resistance mechanisms were observed . However , such changes were not able to elucidate all the complexity observed in the responses of the different sample groups of Aracaju and Crato , related to the evaluated insecticides . In the samples here evaluated , changes in the activity of enzymes related to metabolic resistance were noted , particularly in adult specimens of the vector . However , no difference was observed that was exclusive to the replicates submitted to selection , in both populations . This result suggests the participation of different molecular species and points to the need of high-performance methodological approaches in the identification of resistance mechanisms . However , this type of evaluation is certainly beyond the scope of the tests used in routine resistance monitoring [12 , 43 , 55 , 56] . Regarding Ace , target of OP insecticides , no significant changes were noted neither in total activity ( AChE ) nor in inhibition of activity ( AChI ) . In general , AChI results were concordant for the two classification criteria employed , WHO [40] and Valle et al . [38] . Therefore , the study of the OP target revealed no alterations likely to explain the decrease in malathion susceptibility in the samples exposed to this insecticide . Malathion selection appears to have had a negative impact on deltamethrin resistance , similar to what happened with temephos status of malathion selected mosquitoes . In both cases , temephos and deltamethrin , the RR reduction was more prominent in the malathion selected groups than in the control ones . One potential reason is the selection of malathion resistance mechanisms , in detriment of others , previously present in the original samples and related to resistance to temephos and deltamethrin . In other words , in the absence of changes in Ace , the target of OP , this scenario may result from the deviation of metabolic resources , in the scope of resistance mediated by detoxifying enzymes [57 , 58] . It is worth mentioning that decrease of PY resistance levels in the malathion selected groups was followed by the maintenance ( in Crato ) ou slight reduction ( in Aracaju ) of the kdr resistant homozygotes frequency . Unexpectedly , a marked increase in deltamethrin resistance levels was observed in the control groups , maintained without any insecticide , relative to the original samples , in both populations . This occurred despite a slight decrease in the kdr frequencies , a known mechanism of PY resistance . This decrease in kdr mutations rate is somewhat expected since there are indications that , in the absence of PY , kdr mutations do not offer an adaptive advantage; instead , this feature is related to a significant evolutionary cost [59] . Of course , the possibility of contamination cannot be completely ruled out , although the whole rearing process has been tracked , as well all bioassays steps . However , taken together , such data suggest that ( a ) the kdr mutation frequency assessment should not be used as a single marker of PY resistance , at least in A . aegypti , and ( b ) additional mechanisms capable of conferring PY resistance were selected during the laboratory procedures . In general , selection experiments performed under laboratory conditions tend to privilege multiple low-effect mechanisms , as would be expected in the field , according to assumptions of natural selection [60] . However , in the field , application of insecticides seems to be just one amongst the many challenges faced by mosquitoes . Aedes aegypti , as a generalist insect , can also deal with the presence of pollutants or secondary plant metabolites in its larval breeding sites . This condition could bestow a pre-adaptive advantage regarding the insecticide challenge and justify the early detection of resistance to newly introduced compounds , for instance , in the context of vector control programs [20 , 21] . In fact , some metabolites and pollutants were able to induce a significant increase in the expression of MFO enzymes in A . aegypti , corroborating such pre-adaptability for insecticide resistance [61] . As expected , higher changes in resistance levels were observed in the larval stage , submitted to selection . However , the metabolic resistance components evaluated were more altered in adults . The methodology used here for the quantification of resistance mechanisms is the approach of choice in the routine of the resistance monitoring of vector populations . However , this methodology has known limitations: it uses general substrates , which reveal classes of enzymes and not specific molecular entities . This means that eventual changes in some particular enzyme of a given class can be masked by the pool of enzymes acting on that substrate . In addition , it is possible that other mechanisms , not investigated by this approach , are in operation . Evaluation of metabolic activities in adult Aracaju females indicated a potential participation of GST enzymes in the increased deltamethrin resistance: after comparison with both criteria , parental 99th percentile and the nonparametric statistics , higher GST activity was detected in Aracaju samples kept without insecticide—precisely those with a marked increase of PY resistance levels . Some studies suggest relation between GST and PY resistance of A . aegypti populations from different geographic origins [43 , 62 , 63]; others have even characterized this association [49 , 64] . However , it is worth noting that in adult females from Crato control groups , although also bearing an increased deltamethrin resistance , the change in GST was not prominent , being only detected through medians comparison . It is well known that insecticide resistance has a multifactorial character . There are several recent examples and there is growing understanding that different populations of the same vector species can select different resistance mechanisms against the same class of insecticides . This can derive from several parameters , intrinsic and extrinsic . Among the former are the genetic background and previously selected mechanisms , while insecticide selection pressure intensity and sources are among the extrinsic factors [15 , 46 , 63 , 65 , 66] . Although malathion is an adulticide , we opted to employ larvae in the selection procedure for operational reasons . On one hand , we were interested in carrying out this analysis in a timely manner to respond to the management of resistance in the country; on the other hand however , the technical difficulty of calibrating selection trials with adults is greater than with larvae . Selection with larvae was then a cost-effective way in which we decided to invest in order to ensure that all specimens would be exposed to malathion in a way as controlled as possible . Culicidae males are smaller and develop faster than females . In this case , the ideal selection with adults would require different experiments , with distinct malathion concentrations for males and females . Furthermore , in bioassays with adults , mainly with the available impregnated surfaces , it is difficult to control the level of exposure , since contact with insecticides occurs only during landing . Several studies report the use of larvae in selection experiments with adulticides and , in general , this apparent inconsistency is not even mentioned in the texts ( examples in [67–78] ) . Despite this , we are aware that the selection of larvae with malathion here shown may not represent the field situation reliably , and the results may have limited value in defining resistance management recommendations . For example , there is at least one report of the development of resistance to the adulticide deltamethrin that was more effective in larval selection , compared to adults [79] . In another study , selection of adult females with the same PY also resulted in only mild resistance status alteration of a Brazilian A . aegypti population [80] . It is possible that adaptation to laboratory rearing , in the absence of insecticides , has favored , or selected , one or more mechanisms relevant to the response to PY . Unlike changes in NaV ( kdr ) , such mechanisms could have borrowed advantage to specimens carrying them . This might have occurred even in the absence of selective pressure and under optimal rearing conditions . If this is the case , malathion selection may have hampered the expression of such mechanisms in some way: lower levels of PY resistance were detected in the selected specimens compared to the control groups . In general , it is expected that laboratory rearing , without selective pressure ( as was the case with the control groups herein ) , will increase homogeneity regarding insecticide susceptibility [81] . However , our data indicate that this relationship , between absence of selective pressure and reduction of heterogeneity , although frequent , cannot be generalized . Slope is often a parameter put aside in the insecticide resistance monitoring of disease vectors . Few studies mention the heterogeneity of the evaluated samples , or often fail to inform details ( reports we detected that mention slope results: [27 , 28 , 37 , 56 , 82–88] ) . In the present study , homogeneity increase was noted in most groups , regardless the insecticide challenge , probably due to rearing under laboratory-controlled conditions . The main exception occurred with deltamethrin evaluation: control groups of both populations presented greater heterogeneity related to the PY . And it was precisely for deltamethrin that an unusual and marked resistance increase was detected in these control groups . Laboratory selection of samples from two Brazilian A . aegypti natural populations with the OP malathion was attained . In both cases , alterations in the resistance profiles related to other insecticides employed in the country against A . aegypti were also observed . No cross-resistance was detected between malathion and temephos , also an OP , nor between malathion and deltamethrin , a PY . On one hand , malathion selection had a negative effect on resistance to other insecticides . On the other hand , maintenance of both populations in an insecticide-free environment resulted in an unexpected increase in resistance to deltamethrin . However , the methodological procedures routinely adopted by the control programs did not reveal the potential resistance mechanisms associated , in each case . This scenario suggests the pertinence of adopting high throughput investigation approaches which we are , in fact , providing with the material presented here . Our data point to the feasibility of applying diflubenzuron and malathion in A . aegypti control in Brazil , respectively against larvae and adults . A larvicidal rotation scheme , as already recommended by the MoH , could certainly contribute to preserve diflubenzuron ( or another CSI ) in the field as a viable strategy . Unfortunately , a similar approach with malathion is not possible since all the alternative available adulticides are PY compounds [23] , and A . aegypti resistance against this class of insecticides is widespread throughout the country [20 , 21 , 29 , 44 , 56] . Finally , considering the insecticides employed by control programs , our results reinforce the relevance of investigating the susceptibility status , and the associated mechanisms , of more field A . aegypti populations . This policy could subsidize the rational choice of the compounds and collaborate to keep the feasibility of the few available compounds .
|
Dengue , Zika and chikungunya viruses affect millions of people worldwide . Due to the lack of specific antivirals or to the limited supply of vaccines , focus remains on the control of the main vector , Aedes aegypti . Although the importance of social participation in the elimination of A . aegypti breeding sites is increasingly recognized , chemical control is still an important component of vector control . The exaggerated use of insecticides results in the spread of resistance and , consequently , in the loss of their effectiveness . In Brazil , malathion is the last adulticide available to the control of A . aegypti , due to the widespread resistance to pyrethroids . In order to anticipate what could occur in the field , we exposed two vector populations to selection with malathion . Both malathion and temephos , a larvicide largely employed , are organophosphates; however , they are structurally distinct molecules and seem to elicit different resistance mechanisms . We confirmed this issue: selection with malathion had a negative impact on temephos resistance compared to groups reared without any insecticide . Indeed , the variety of responses of both vector populations to the various insecticides points to the participation of multiple resistance mechanisms and confirms previous assumptions regarding the difficulty of identifying diagnostic insecticide resistance mechanisms .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"geographical",
"locations",
"microbiology",
"malathion",
"organic",
"compounds",
"animals",
"organophosphates",
"developmental",
"biology",
"pharmacology",
"insect",
"vectors",
"agrochemicals",
"infectious",
"diseases",
"antimicrobial",
"resistance",
"south",
"america",
"aedes",
"aegypti",
"life",
"cycles",
"chemistry",
"disease",
"vectors",
"insects",
"agriculture",
"brazil",
"arthropoda",
"insecticides",
"people",
"and",
"places",
"mosquitoes",
"eukaryota",
"organic",
"chemistry",
"ethers",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"species",
"interactions",
"larvae",
"organisms"
] |
2018
|
Laboratory selection of Aedes aegypti field populations with the organophosphate malathion: Negative impacts on resistance to deltamethrin and to the organophosphate temephos
|
The mechanical properties of virus capsids correlate with local conformational dynamics in the capsid structure . They also reflect the required stability needed to withstand high internal pressures generated upon genome loading and contribute to the success of important events in viral infectivity , such as capsid maturation , genome uncoating and receptor binding . The mechanical properties of biological nanoparticles are often determined from monitoring their dynamic deformations in Atomic Force Microscopy nanoindentation experiments; but a comprehensive theory describing the full range of observed deformation behaviors has not previously been described . We present a new theory for modeling dynamic deformations of biological nanoparticles , which considers the non-linear Hertzian deformation , resulting from an indenter-particle physical contact , and the bending of curved elements ( beams ) modeling the particle structure . The beams’ deformation beyond the critical point triggers a dynamic transition of the particle to the collapsed state . This extreme event is accompanied by a catastrophic force drop as observed in the experimental or simulated force ( F ) -deformation ( X ) spectra . The theory interprets fine features of the spectra , including the nonlinear components of the FX-curves , in terms of the Young’s moduli for Hertzian and bending deformations , and the structural damage dependent beams’ survival probability , in terms of the maximum strength and the cooperativity parameter . The theory is exemplified by successfully describing the deformation dynamics of natural nanoparticles through comparing theoretical curves with experimental force-deformation spectra for several virus particles . This approach provides a comprehensive description of the dynamic structural transitions in biological and artificial nanoparticles , which is essential for their optimal use in nanotechnology and nanomedicine applications .
Single-molecule techniques , such as Atomic Force Microscopy ( AFM ) , have become widely available to explore the physical properties of biological assemblies . These techniques have triggered extensive research efforts to explore the protein shells of plant and animal viruses , and bacteriophages . A wide spectrum of viruses infect their animal and plant hosts . To do so , these pathogens employ a diverse range of infection mechanisms . Typically , animal cells utilize a mechanism involving molecular recognition of the host via specific cell surface receptors . On the other hand , plant viruses often lack this host infectivity mechanism . In contrast to animal cells , plant cells are enclosed within a rigid cell wall and cuticle . These features represent significant physical barriers to viral infection . Thus , it is thought that physical damage to a plant’s surface that exposes the underlying cells , often through mechanical stress or as a result of insect feeding , is a requirement for viral infection to occur . Regardless of these viral infectivity mechanism differences , dynamic structural transitions by the viral capsids appear to be general features of most viruses . These infectivity mechanism differences are very likely correlated with variable virus capsids’ structures , dynamics and energetics-based mechanical properties . Therefore , understanding viral infectivity represents a clear motivation for investigating capsids’ mechanical and dynamic property differences . The AFM-based mechanical testing of viral nanoparticles has now become the principal tool to probe the physico-chemical and materials properties of viruses [1] . In these experiments , an indenter ( cantilever tip ) approaches a particle and gradually deforms the particle , while the restoring ( indentation ) force F from the particle , corresponding to the particle deformation X , is measured . A variety of viruses have been characterized by profiling F as a function of X ( FX-curve ) , including bacteriophages Φ29 and HK97 [2–4] , the human viruses Noro Virus , Hepatitis B Virus , Human Immuno Deficiency Virus ( HIV ) , Adenovirus ( AdV ) and Herpes Simplex Virus [5–9] , and other eukaryotic cell infecting viruses such as Minute Virus of Mice , Triatoma Virus ( TrV ) and plant viruses Cowpea Chlorotic Mottle Virus ( CCMV ) and Brome Mosaic Virus ( BMV ) [10–14] . The FX-curves reveal valuable information about the particle spring constant , reversibility of deformation , and forces required to deform or distort capsid structures tested mechanically . AFM experiments reveal a surprising diversity of mechanical properties of biological particles . These properties have been shown to correlate with local conformational dynamics of the capsid structure and to contribute to events such as receptor binding , genome uncoating and capsid maturation , all crucial steps in different viral infectious cycles . The main impediment to gaining further energetic and structural insights into these properties is that experiments reveal results that are difficult to interpret without a comprehensive theoretical modeling framework that describes the full range of observed mechanical behaviours . For example , it is not clear why is the initial portion of the FX spectra is weakly non-linear ? Why do the FX spectra for some particles exhibit sudden drops in the deformation force , whereas the FX curves for other particles show gradual force decreases ? What features determine the mechanical limits of the particle , i . e . the critical forces and critical deformations ? Why do the FX spectra differ from one measurement to another for the same particle , even when it is indented along the same symmetry axis ? The latter property points to the stochastic nature of deformation and collapse transitions , but what defines the likelihood of structural collapse at a given force load ? Virus particles are often characterized by their spring constants , but our in silico nanoindentation studies show that the derivative , dF/dX , fluctuates significantly with X [15] . What is the extent of structure remodeling that gives rise to a non-monotonic behavior for dF/dX ? What types of mechanical excitations corresponding to these structure alterations contribute to the particle deformation ? These questions clearly show the need for a thorough theoretical framework describing capsid and other types of nanoshell deformations . A number of theoretical approaches have been designed to describe the dynamics of virus particles , including: finite element analysis [16] , normal mode analysis [17] , elastic network modeling [18] , atomistic MD and coarse-grained simulations [19–22] , and other approaches [23] . Building upon the results from direct MD simulations of mechanical deformation , here we take a step further to develop a systematic approach for meaningful interpretation of the force-deformation spectral lineshapes available from single-particle nanomanipulation experiments . In these state-of-the-art experiments , a slowly moving cantilever tip gradually deforms a biological particle , and multiple nanoindentations are performed to directly probe the particle’s mechanical response . Using slow indenter velocities is entirely justified biologically . This view can be seen to align with the kinetics of genome packaging and ejection , which occur on a second timescale or shorter , as does the associated pressure change occurring inside the particle . For these reasons , we formulate a theoretical model for a uniaxial particle’s deformation achieved using slow indenter velocities . The theory links the slope , critical force , and the critical deformation of the FX-curve with the physical characteristics of the structure , geometry and overall shape of the particle and indenter . First , we summarize the results of Molecular Dynamics ( MD ) simulations of mechanical deformation accelerated on Graphics Processing Units ( GPUs ) [24 , 25] , which we refer to as nanoindentation in silico , of the empty CCMV capsid particle; see S1 Fig [15] . The in-depth analysis of the structure and energy output from MD simulations for this specific example of a thick-shelled nanoparticle has enabled us to identify the most important types of mechanical excitations that contribute to the deformation of biological particles . Next , we formulate the model by analyzing structural evidence from in silico nanoindentation measurements , which mimic the nanoindentation experiments in vitro . Finally , we apply the model to characterize the experimental and simulated FX-spectra for several specific examples of biological nanoparticles: the protein shells of the viruses CCMV , AdV and TrV .
We employed the methodology of “nanoindentation in silico” ( i . e . computational-based indentation of a nanoparticle; see S2 Fig ) [15 , 25 , 26] , which mimics the AFM-based force measurements in vitro . In this approach , the mechanical loading of a biological particle is performed computationally ( Methods and Models section ) using MD simulations with experimental conditions of dynamic force application f ( t ) = rf t . The significant computational acceleration available on Graphics Processing Units ( GPUs ) enables us to apply the experimentally relevant force-loading rates rf = κνf ( κ is the cantilever spring constant ) , which correspond to the cantilever base velocity νf = 0 . 1–1 . 0 μm/s . Structural transitions can be resolved by examining the coordinates of amino acid residues , and biomechanical characteristics can be accessed through analysis of the energy output . Our in silico experiment provides the complete high resolution simulation view of particle deformation and collapse described below , where the choice of simulation conditions is entirely under the control of the investigator . The full control over the system during the nanomanipulations in silico can be used to study deformation at different specific symmetry points on the particle surface as well as the particle-indenter contact area dependence , and to relate the force and energy values recorded at any point in the simulation to the specific details observed in the particle’s structure . This type of precise high resolution control is not possible from nanoindentation carried out experimentally . Furthermore , when a sufficiently slow force loading is utilized our approach to nanoindentation in silico allows the investigator to follow the stochastic dynamics of mechanical deformation of a biological particle , which is microscopically reversible . In this regime of compressive force application , the rate of force increase is slower than the rate of system re-equilibration at each point along the deformation reaction path ( quasi-equilibrium ) . For these reasons , we utilize our nanomanipulations in silico to guide the detailed modeling and interpretation of experimental results for the deformation dynamics of any biological nanoparticle being studied . In this section , we will utilize MD simulation data to motivate the FNS model . Rigorous analyses of the structures and energy outputs from MD simulations of mechanical deformation of viruses [15] and a microtubule [26] showed that the mechanical response of biological nanoparticles , subject to a uniaxial deformation , can be divided into Hertzian and bending contributions ( see Fig 1 ) . In this section , we exemplify the FNS model in terms of the mechanical degrees of freedom that we identify to be most relevant to the uniaxial type of deformation . In dynamic force-ramp f ( t ) = κνf t , an indenter ( cantilever tip ) compresses a particle ( Fig 1 and S2 Fig ) , thus creating a physical contact between them . The force loads the particle mechanically over time t with the force-loading rate κνf ( κ and νf are the respective cantilever spring constant and velocity ) . For small force , the mechanical energy is localized to the particle surface under the tip , and the tip and particle undergo normal displacements utip and upar , corresponding to the deformation xH = utip+upar . Since utip≪upar , xH = upar . The force gradually loads the particle , stressing the side portions of the structure undergoing bending deformations xb ( Fig 1 ) . Force-ramp conditions project the complex dynamics of the particle deformation in the direction perpendicular to the particle surface . During nanomanipulations in vitro and in silico , the deformation force F , the mechanical response of the particle , is measured as a function of the total deformation X = xH+xb ( reaction coordinate ) . Therefore , we focus here on the computation of the force-deformation ( or FX ) lineshape . We quantified xH and xb directly using the simulation output for the CCMV particle and found these to be independent , small-amplitude deformations . For example , the maximum values of xH and xb for the CCMV shell are 3 nm and 4 . 3 nm , respectively ( Fig 1b and 1e ) . Analysis of the experimental FX-spectra and structure snapshots from the MD simulations showed that the Hertz model [29 , 30] properly accounts for the force FH due to the observed local curvature change of the particle under the tip ( Fig 1a ) , F H ( x H ) = 1 D H R p a r R t i p R p a r + R t i p · x H 3 / 2 ( 4 ) where Rpar and Rtip are the radii of the particle and the tip , respectively . The term DH is given by D H = 3 4 1 - σ H 2 E H + 1 - σ t i p 2 E t i p ( 5 ) where EH and Etip are the Young’s moduli and σH and σtip are the Poisson’s ratios for the particle and the tip , respectively . Since Etip≫EH , D H = 0 . 75 ( 1 - σ H 2 ) / E H . To describe the bending deformations Fb ( xb ) , we discretize the side portion of the particle structure ( barrel ) into curved vertical beams of length L ( Fig 1d ) . The results of comparison of the out-of-plane bending and the in-plane stretching modes of deformation ( Fig 3 ) showed that the effect of in-plane stretching on the total particle deformation is indeed negligible ( see previous section ) . Hence , we can safely assume that the length of vertical beams L does not change with total deformation X . In view of the observations described above , our discretization of the particle barrel into curved vertical beams is fully justified . For a spherical particle of thickness r , the total number of beams is N = 2πRpar/r . For small beam deformation xb ( Fig 1 ) , the potential energy change is given by the integral Eb I/2∫L ( κ ( xb , l ) −κ0 ) 2 dl [29 , 31] , where κ0 and κ ( xb , l ) are the initial and instantaneous beam curvatures ( 0≤l≤Rpar−xb/2 ) and Eb I is its flexural rigidity , given by the product of the Young’s modulus for bending Eb and the moment of inertia I . With the beam shape function q ( x b , l ) = ( R p a r + x b 2 ) 1 - l 2 ( R p a r - x b / 2 ) 2 ( 6 ) the curvature is given by κ ( x b , l ) = q ″ ( x b , l ) ( 1 + ( q ′ ( x b , l ) ) 2 ) 3 / 2 ( 7 ) where q′ and q″ are the first and second derivatives of q with respect to l . By performing the integration we obtain the expression for the bending energy , which upon differentiation with respect to xb , gives the bending force . Expanding the resulting expression in Taylor series in powers of xb and retaining the linear term in the expansion , we obtain: f b ( x b ) ≅ 9 E b I π 8 R p a r 3 · x b ( 8 ) Combining the contributions from all N coupled elements ( beams ) and adding Eqs ( 4 ) and ( 8 ) together , we obtain the deformation force F ˜ ( x H , x b ) = k H x H 3 / 2 + N k b x b , where kH = ( Rpar Rtip/ ( Rpar+Rtip ) ) 1/2/DH is the “Hertzian spring constant” and k b = 9 E b I π / ( 8 R p a r 3 ) is the beam spring constant . In agreement with in silico indentations of CCMV shell ( Figs 2 and 4 ) and recent experiments on thick-shelled particles [11] , F ˜ predicts that the initial portion of FX-curves is weakly nonlinear , but fails to capture the force drop ( Fig 4 ) because the theory lacks a description of structural damage ( see next section ) . As we pointed out , owing to the discrete arrangements of capsomers forming the CCMV shell , structural elements fail but not all at the same time . To reflect the discrete nature of microscopic transitions , we represent a particle by a collection of N identical coupled elements ( beams ) interacting with an indenter through a Hertzian cushion ( Fig 1 ) . Each i-th beam undergoes the elastic deformation xbi = xb with the spring constant kb until it fails mechanically when the load on the beam reaches some critical value f b i * ( see snapshots in Fig 4 ) . The spherical geometry of a virus particle dictates the parallel arrangement with the spring K b = ∑ i = 1 N k b i = N k b . At any given time , there are n ( or N−n ) beams that have failed ( or survived ) , and the actual bending force is given by F b ( x b ) = k b ( N - n ) x b = K b x b ( 1 - n N ) ( 9 ) We define the probability of damage d = n/N and survival s = ( N−n ) /N = 1−d of the collection of beams , which in the continuous limit are described by the probability density function ( pdf ) g ( Fb ) , i . e . d ( F b ) = P r o b ( F b ) = ∫ 0 F b g ( F b ′ ) d F b ′ ( 10 ) and s ( Fb ) = 1−d ( Fb ) ( Fb = Kb xb ) . With the structural damage accounted for , the deformation force becomes: F ( x H , x b ) = k H x H 3 / 2 + K b x b s ( x b ) ( 11 ) Our rationale for using the survival probability measure s ( xb ) is based on the in-depth analysis of structures from the MD simulations of mechanical deformation of the CCMV virus [15] and microtubule polymers [26] . Both systems clearly demonstrate that soft biological particles accumulate structural damage . Fig 3 shows formation of small cracks in the CCMV shell , whereas Fig 1e provides a global view of the extent of structural damage in the CCMV particle , accumulated in the course of deformation . The transition to the collapsed state occurs when all beams have failed , and so , the longest lasting beam determines the collapse onset at the critical deformation Xcol when tension exceeds the critical force Fcol . Hence , the statistics of the maximum ( extreme ) force determines the beams’ failure . For these reasons , we used the two-parameter Weibull distribution [32] s ( x b ) = exp - F b F b * m = exp - K b x b F b * m ( 12 ) with the cooperativity parameter m , and the scale parameter F b * . The meaning of F b * can be understood by using the condition of maximum force , dFb/dxb = 0 , from which we obtain: F b * = K b x b * m m where x b * is the critical beam deformation . By substituting F b * into the expression for Fb ( xb ) , we obtain the bending force threshold F b c o l = F b * e m m = K b x b * e m ( 13 ) Finally , by substituting Eq ( 12 ) for s ( xb ) in Eq ( 11 ) , we obtain one of the main results of the paper: F ( x H , x b ) = k H x H 3 / 2 + K b x b exp - K b x b F b * m ( 14 ) The Fluctuating Nonlinear Spring ( FNS ) model describes the nonlinear deformation as a superposition of the weakly nonlinear deformation ( Hertzian cushion ) and the elastic deformation ( particle barrel; Fig 1 ) of varying stiffness that is gradually degraded with X . Consequently , Eq ( 14 ) shows that the uniaxial deformation and structural collapse of a biological particle can be represented by the mechanical evolution of a fluctuating weakly nonlinear spring . This behavior led us to propose the name of the model . The beams’ bending starts as elastic ( Nkb ) , but becomes increasingly more stochastic near the collapse transition , thus explaining the variability of Fcol and Xcol in the experimental and simulated FX-spectra ( see Figs 4 and 5; see also S3 and S4 Figs ) .
The FNS model can be understood by adopting a picture of “a particle as barrel” under the Hertzian cushion . Then , Eq ( 14 ) can be viewed as the complex mechanical response function to describe a biological particle whose stiffness is degraded exponentially with xb as K b exp [ - ( K b x b / F b * ) m ] . The latter quantity can be taken as the “effective stiffness” as compared to the native state stiffness of the particle Kb . If we multiply the number of beams N = 2πRpar/r by kb ( given by the prefactor in Eq ( 8 ) ) , we obtain K b = N k b = 9 E b I π 2 / 4 R p a r 2 r , which carries information about the “particle barrel” ( no information about beams ) . Also , in Eq ( 14 ) , the shape parameter m can be interpreted as a cooperativity parameter that takes into account dynamic coupling among the beams . When m = 1 the beams are independent , which corresponds to the exponential distribution for the survival probability s ( x b ) = exp [ - K b x b / F b * ] . When m ≠ 1 , the structural elements behave cooperatively to withstand the stress . In the FNS model , the main quantity F is a bivariate function ( Eq ( 14 ) ) , but in the experiment F is measured as a function of the sum X = xH+xb . To resolve xH and xb for each value of X we use the following considerations . A particular realization of the deformation process ( FX-trajectory ) is a stochastic path on the 2D surface F ( xH , xb ) displayed in Fig 5 . For slow loading , when the particle structure equilibrates on a timescale faster than the rate of force change , the dominant path is the equilibrium deformation path . Utilizing slow cantilever velocities ( νf = 0 . 06–1 . 0 μm/s ) allows us to use this quasi-equilibrium argument . Importantly , our recent study of the dynamics of deformation and collapse of microtubule polymers [26] show that in silico indentation experiments reported here are carried out under near-equilibrium conditions of compressive force application . Then , the equilibrium force can be determined from the requirement that the deformation force ( and deformation energy ) attains the minimum . Hence , finding the minimum force for each value of X is equivalent to finding xH and xb , which minimize F ( xH , xb ) subject to the constraint , X = xH+xb . This can be solved using the method of Lagrange multipliers summarized in the S2 Text . The average simulated spectra for the CCMV particle are compared with the theoretical curves in Fig 4 ( simulated spectra for CCMV are accumulated in S3a Fig ) . To find the best fit , we employed two methods . The exact method is based on Eq ( 14 ) and uses Lagrange multipliers to find xH and xb subject to the constraint X = xH+xb . This approach can be used to model the average force-deformation spectra . The application of this method to describing the experimental or simulated force-deformation spectra requires solving the nonlinear equation for beam-bending deformation xb: a 1 x b 4 m + a 2 x b 3 m + a 3 x b 2 m + a 4 x b m + a 5 x b + a 6 = 0 , where a 1 = m 2 K b 2 ( K b / F b * ) 4 m , a 2 = - 2 m ( 1 + m ) K b 2 ( K b / F b * ) 3 m , a 3 = ( 1 + 4 m + m 2 ) K b 2 ( K b / F b * ) 2 m , a 4 = - 2 ( 1 + m ) K b 2 ( K b / F b * ) m , a 5 = 9 / 4 k H 2 , and a 6 = K b 2 - 9 / 4 k H 2 X are constant coefficients . Then , xH is obtained as xH = X−xb . In the piece-wise approximate method ( see inset in Fig 4 ) , a spectrum is divided into the Hertzian-deformation-dominated initial regime I: X≈xH ( xb ≈ 0 ) and F≈FH = kH X3/2; and the transition regime II ( corresponding to the non-monotonic part of the FX-curve ) : X≈xb and F≈Fb = Kb Xs ( X ) . We calculate FH in regime I for X ≈ x H < x H m a x , where x H m a x is obtained using Lagrange multipliers and setting s ( xb ) = 1 . In regime II , we use F ( x H m a x , x b ) = k H ( x H m a x ) 3 / 2 + K b ( x b - x H m a x ) exp [ - ( K b ( x b - x H m a x ) / F b * ) m ] for X ≈ x b > x H m a x . This method can be used to model individual FX-spectra ( displayed in S3 and S4 Figs ) in order to access the entire distributions of a particle’s mechanical and statistical characteristics and to probe the variability of these properties due to the intrinsically stochastic nature of mechanical deformation and collapse of biological particles . We applied the FNS model-based theory to describe FX curves for the CCMV particle . The agreement between the simulated force-deformation spectra and theoretical FX-curves for the CCMV particle is very good ( Fig 4 ) . The FX-spectra presented in Fig 4 also fully agree with the FX-spectra for the CCMV particle discussed extensively in our previous study [15] in terms of the critical force Fcol , critical deformation Xcol , and the slope dF/dX . Simulated FX-curves show smaller variability as compared to the experimental FX-spectra , because in experiments not only are 2-fold , 3-fold , and 5-fold icosahedral orientations probed , but also various intermediate orientations . Less sharp force peaks due to slower force decrease observed in simulations can be attributed to overstabilizing the inter-chain interactions and neglecting the hydrodynamic interactions in the SOP model of the CCMV shell ( work in progress ) . The values of model parameters obtained using both methods of estimation of the contributions xH and xb are very close ( Table 1 ) . For all symmetry types , the Hertzian excitation is softer than the bending ( kH < Kb ) , implying smaller Young’s modulus , EH < Eb , which is why the Hertzian degree of freedom is excited first ( regime I; see Fig 1 ) . After the Hertzian force reached the maximum F H m a x = k H ( x H m a x ) 3 / 2 at X ≈ x H = x H m a x , a subsequent force increase excites the beam-bending degrees of freedom ( regime II ) and xH ( xb ) decreases ( increases ) ; see Fig 2a . Hence , the physical properties of the particle are dynamic ( rather than static ) since the nature of its mechanical response changes with increasing X from Hertzian-type to beam-bending deformation . The gradual decrease in xH is somewhat counter-intuitive as one expects that xH ( and FH ) remains constant after it has reached the maximum x H m a x ( and F H m a x ) . This is because the actual stiffness of beams is not constant but is degraded with increasing xb due to the consecutive beam failure events ( and accumulated damage ) . Therefore , in the transition regime II , the beam-bending xb increases not only due to the continuing mechanical loading , but also as a result of stress redistribution to intact beams . The FNS model also explains why the mechanical response of biological particles depends on the structure of the particle-indenter contact and the particle and indenter geometries [15] . The parameter obtained from the model for different symmetries show that the mechanical response of CCMV varies with the location of compressive force application ( Table 1 ) . As all virus shells reflect the discrete symmetry of their specific capsomer arrangements , these results imply that the mechanical properties of virus particles are local ( i . e . location-specific ) characteristics of their structure . Furthermore , we found in our previous studies of near-spherical virus particles [15] and cylinder-shaped microtubule polymers [26] that the deformation force depends on the indenter size . The FNS model fully accounts for this finding . In the FNS model , the information about the particle and indenter geometries is contained in the Hertzian spring constant kH . Hence , the model predicts that the geometric effects are important only in the initial Hertzian-deformation dominated regime ( regime I; see Fig 4 ) . Application of the FNS model to several nanoscale biological particles ( CCMV , AdV , and TrV virus shells ) revealed that all exhibit m>1 , which means that the structural elements forming the side-portion of the biological particle structure are mechanically coupled . For example , for the CCMV particle we found that for all indentation locations , the range of values was 1 . 8<m<2 . 1 ( Table 1 ) . Therefore , positive cooperativity is exhibited by the side-portion of the particle’s structure ( beams ) , regardless of the point of indentation . Interestingly , the beams do not just fail when F>Fb* , but begin to fail under smaller force F b c o l = F b * / e m m . For example , for m ≈ 2 , we obtain F b c o l ≈ 0 . 43 F b * . The AFM-based measurements for the empty CCMV shell , empty TrV capsid , full TrV virion ( with encapsulated ssRNA molecule ) and full AdV virion ( with encapsulated dsDNA ) are presented in S3b and S4 Figs . Theoretical fits to the experimental average FX-curves shows that their deformations are well described by the FNS model ( Fig 5 ) . The obtained Young’s moduli for Hertzian deformation are uniformly smaller ( ∼10–100 MPa ) than the Young’s moduli for bending deformation ( Giga-Pascal range; Table 1 ) . There are small variations in the model parameters for the AdV virion due to force application at locations with different symmetry axes . This correlates with our similar findings for the CCMV shell , implying that the symmetry of local arrangements of capsomer repeats at the point of indentation influences its mechanics [15] . The values of cooperativity parameter are found to be greater than unity ( m > 1 ) , representing positive cooperativity , for all the systems studied . Parameters for empty and ssRNA-loaded TrV capsids indicate that the difference in particle stiffness is largely due to an increase in the Young’s modulus for Hertzian deformation EH = 0 . 03 GPa ( empty TrV ) vs 0 . 14 GPa ( full TrV ) , which suggests that local indentations are resisted in ssRNA-filled particles . These results fit with the previously observed deformation of RNA-filled TrV into an oblate sphere to maximize the volume available to pack the genome [12] . Hence , confining the large ssRNA genome inside the small particle volume builds internal pressure resisting local indentation . This behavior is in agreement with the general property of bacterial and higher organism viruses that have evolved to achieve maximum nucleic acid packing into the available virion volume , often exhibiting significant internal pressures in the mature packaged state . It is known that genomic material is one of many factors that influence nanoparticles’ mechanics , as described , in one example , in our previous study of TrV [12] . In full accord with this notion , the FNS model predicts that the presence of the genome defines the stability and physical properties of native virus particles . The biochemical properties of the nanoparticle shell are defined by the intra- and intersubunit protein interactions , and these non-covalent interactions are fully reflected in the SOP-model and they show up in the simulated FX-curves . Previously , the 3D Young’s modulus of the capsid material was estimated by investigators using a thin shell theory [1 , 11 , 12 , 29] . This assumption is valid for some bacteriophage capsids , but is not so in the case of CCMV and TrV capsids where the shell thickness cannot be neglected with respect to the virion radius . The FNS model properly accounts for compression of the protein layer under the tip . In the FNS model , the beam-bending modulus ( Eb ) is roughly equivalent to the 3D Young’s modulus in the thin shell theory . It is estimated at ∼0 . 85 GPa ( experiment ) and ∼0 . 4–0 . 5 GPa ( simulations ) for the empty CCMV capsid ( Table 1 ) . These are similar to yet larger than the values of 0 . 15–0 . 30 GPa obtained with thin shell theory [1 , 11] and 0 . 28–0 . 36 GPa from finite-element analysis ( ∼0 . 25 GPa ) [33] , but they disagree with the estimates from several computer modeling studies ( 0 . 08–0 . 09 GPa ) [22 , 23] . In the modeling study based on spherical harmonics [23] , multiple deformation modes have also been observed , corresponding to equilibrium deformations of the polar regions ( tip-surface contact area in FNS model ) and the side wall ( beams in FNS model ) of the shell . For the empty TrV capsid , we obtain Eb ≈ 0 . 9 GPa ( Table 1 ) whereas the thin shell theory gives ∼0 . 5 GPa . The lower previous estimates of the 3D Young’s modulus result from attributing the softer Hertzian deformation mode to bending of the capsid shell in the thin shell theory . Indeed , for CCMV and TrV , the thin shell theory estimates of 0 . 15–0 . 30 GPa and 0 . 5 GPa are between the values of EH = 0 . 02–0 . 03 GPa and Eb = 0 . 85–0 . 95 GPa from the FNS-model based modeling ( Table 1 ) . One of the novel aspects of the FNS model is that it allows one to interpret the survival probability s ( x b ) = e x p [ - ( K b x b / F b * ) m ] ( Eq ( 12 ) ) using the concept of structural similarity quantified by the structure overlap function χ ( xb ) . In silico , s ( xb ) can be directly accessed by calculating the structural similarity χ between a given ( current ) structure ( corresponding to beam-bending deformation xb ) and the native ( reference ) state , using the formula: χ ( x b ) = ( 2 M ( M - 1 ) ) - 1 ∑ Θ ( | r i j ( x b ) - r i j ( 0 ) | - β r i j ( 0 ) ) ( 15 ) In Eq ( 15 ) , M is the total number of amino acid residues comprising the particle’s structure ( system size ) , and in the Heaviside step function Θ ( x ) , defined as Θ = 0 for x<0 and Θ = 1 for x≥0 , rij ( xb ) and rij ( 0 ) are the distances between the i-th and j-th amino acids in the given and native structures , respectively ( β = 0 . 2 is the tolerance for distance change ) . Since Hertzian deformation is local , i . e . it is limited only to the protein domains in and around the indenter-particle contact area ( see Fig 1a–1c ) , this type of deformation does not significantly affect the global particle structure , and so χ ( X ) ≈χ ( xb ) . Indeed , the X-dependent profiles of χ show that the structure overlap decreases at large values of X only when mechanical loading starts deforming the beams ( Fig 2b ) . On the other hand , s ( xb ) decreases only after the Hertzian deformation has reached the maximum x H = x H m a x . At this point , a subsequent increase in X loads the beams , resulting in the increase of xb and decrease of s ( xb ) ( Fig 2a ) . Hence , the dependence of s on X can be approximately described as s ( X ) ≈ s 0 Θ ( x H m a x - X ) + s ( X - x H m a x ) Θ ( X - x H m a x ) ( 16 ) In Eq ( 16 ) , s0 = 1 represents the initial values of s ( X ) in the Hertzian deformation regime I , and the second term on the right describes the dependence of s ( X ) in the beam-bending regime II ( see Fig 2 and the inset to Fig 4 ) . Because the structure overlap χ ranges from χ = 1 ( identical structures ) to χ = 0 ( completely dissimilar structures ) , structural alterations and , hence , changes in χ can be translated to changes in s , i . e . s ( X ) ≈ χ ( X ) ( 17 ) Therefore , as Eq ( 17 ) implies , the survival probability s ( X ) can also be modeled using the structure data from nanoindentation simulations . To confirm the above conclusion , we estimated s ( X ) using the structure output from in silico nanoindentations of the CCMV particle . The structure overlap χ-based estimation of s ( X ) ( data points; Eq ( 17 ) ) and theoretical profiles of s ( X ) ( curves; Eq ( 16 ) ) are directly compared in Fig 2b . The results of comparison fully confirm this conclusion , and also demonstrate that the survival probability s ( X ) has a well-defined interpretation in terms of the particle’s structure . Stated differently , this probability measure is not some intermediate variable used to formulate the theory , but rather , it is an important ingredient of the FNS model . Hence , in the theoretical framework of the FNS model , the survival probability s ( X ) provides a direct link between the dynamic structural changes observed in biological particles and the intrinsically stochastic nature of their deformation and transition to the collapsed state . In a sense , the FNS is also a structure-based model . In this regard , the structure overlap function χ ( X ) can be utilized in conjunction with the structure output from nanoindentations in silico to guide the modeling efforts in order to resolve s ( X ) . The proof of a theory is in its predictive power . First , we used parameters of the FNS model ( Table 1 ) to calculate the position Xcol and the amplitude of the force peak ( force maximum ) Fcol for the average FX-spectra ( Figs 4 and 5 ) , and to predict the critical force for collapse: X c o l = x H * + x b * and F c o l = F H ( x H * ) + F b c o l ( x b * ) = k H ( x H * ) 3 / 2 + K b x b * e m ( 18 ) Remarkably , the obtained theoretical values of Fcol ( Table 1 ) agree well with their counterparts extracted from the average FX-curves , which validates the model . Second , individual FX-curves display large variability of critical deformations and critical forces ( see S3b and S4 Figs ) . In the FNS model , this information is implicitly contained in the survival probability s ( xb ) and damage probability d ( xb ) . The width of the transition region , in which s ( xb ) ( d ( xb ) ) decrease ( increase ) to zero ( unity ) , defines the range of critical deformations ΔXcol . Upon rescaling , Kb xb→Fb , s ( xb ) and d ( xb ) are transformed into the force probabilities s ( Fb ) and d ( Fb ) , and the width of the transition region for s ( Fb ) and d ( Fb ) defines the range of critical forces ΔFcol . As an example , we estimated ΔXcol by analyzing the transition range for the survival probability s ( X ) given by Eq ( 16 ) . We used the FNS model parameters obtained for experimentally tested empty CCMV particle from Table 1 ( Fig 5a; see also S3b Fig for individual FX-curves ) . The results of estimation of the transition range for the CCMV shell using s ( X ) are displayed in S5 Fig . We obtained ΔX ≈ 8 . 0 nm ( shaded area in S5 Fig ) which compares well with the experimental value ΔXcol = 6 nm . The corresponding range of critical forces , ΔF = KbΔxb = KbΔX = 0 . 24 nN/nm × 8 nm ≈ 1 . 9 nN , compares well with the experimental range ΔFcol = 0 . 7 nN . Clearly , the experimental ranges for both ΔXcol and ΔFcol are shorter than the theoretical widths ΔX and ΔF due to a limited number of experimental measurements ( 7 runs; see S3b Fig ) . We have demonstrated that the FNS model based theory: ( i ) correctly predicts the location of the force peaks Xcol and amplitude of peak forces Fcol extracted from the average FX-spectra , and ( ii ) describes the variability of critical deformations and critical forces around their average values ( Xcol and Fcol ) . About half of all known viruses possess icosahedral symmetry [34] and , therefore , here we focused on examples of virus particles with this symmetry . However , the model can be applied much more widely to characterize a range of biological nanoparticles , for which the FX-spectra are already available , including plant and animal viruses and bacteriophage , cellular nanocompartments , cytoskeletal polymers , etc . Although the FNS model is tailored to treat small deformations , it can be used to account for large deformations as well . This would require the extension of Eq ( 8 ) to include the higher order terms in xb . Also , the Hertz model could be improved to account for the non-local deformations . The FNS model can be used to interpret the FX-curves for biological particles of different regular geometries , including cylindrical or ellipsoidal shapes , as long as the particles are subjected to a uniaxial compressive force induced by a spherical-like indenter . Extension of the FNS model to other indenter geometries is also possible . In this paper , however , we used the “Hertzian spring constant” kH to treat the sphere-sphere interaction , because our goal was to explore the mechanical deformation of virus particles , which are nearly spherically-shaped , and because the cantilever tips used in AFM experiments can be approximated by a sphere . Also , when nanoindentation measurements are performed using a smaller tip compared to the size of the biological particle ( which is a typical situation realized in AFM experiments ) , the tip-particle contact area is roughly circular . For these reasons , in this paper we treated the simplest case of near-circular contact area . We will discuss these geometry-related aspects of the FNS model in future work , including a more general case of elliptic particle-indenter contact area ( manuscript in preparation ) . Living organisms have evolved with hierarchical supramolecular systems playing key roles in their biological functions . The dynamic properties of spontaneous assembly , disassembly , and self-repair exhibited by supramolecular assemblies explains their central importance . Prime examples of hierarchical supramolecular assemblies are the easily studied plant and animal viruses and bacteriophages . Although well studied , it remains a challenge to elucidate the structural origins of their unique physico-chemical properties as well as to resolve the specific mechanisms of their response to a wide variety of both biochemical molecules and external mechanical factors . In conjunction with single-molecule techniques , like AFM , dynamic force spectroscopy has become a nearly routine discovery tool for understanding the physical properties of intact biological particles [1] . However , the results of such experimentation remain difficult to interpret . In a number of our recent studies , we have developed an approach to nanoindentation in silico that involves multiscale modeling [15 , 26] . The value in this novel approach is that it provides a toolbox for the computational interrogation of biomechanical properties that characterize large-size biological assemblies . As a result of this recent progress in experimental and computational studies on forced indentation of biological nanoparticles , there is a growing need for a simple theoretical approach to quantitatively describe force-deformation curves . We developed the analytically tractable FNS model which uniquely combines the elements of continuum mechanics and statistics of extremes to accurately describe the uniaxial mechanical deformation and structural collapse ( beyond buckling ) in biological nanoparticles . The FNS model is based on a clear microscopic picture resulting from the multiscale modeling efforts , which involve direct atomistic and coarse-grained simulations of virus particles . However , it is important to note an application of the FNS model does not require the results of MD simulations as an input , and , hence , the FNS model can be applied widely to any regular geometry nanoparticle . To formulate the model , here we used: ( i ) virus deformation simulation data which agreed with experiment [15] , ( ii ) data gathered at the nanometer scale ( <1 nm ) , and ( iii ) experimentally relevant force-loading conditions . Due to the limited resolution of the AFM-based experimental technique , the only direct structural evidence is currently available from in silico experiments ( Figs 1 and 2 ) , which we have utilized in this paper to guide our modeling efforts . We have demonstrated how the FNS theory can accurately model the deformation of viral nanoparticles , showing promising applications of this theory to describing the physics and mechanochemistry of a wide variety of both natural as well as synthetic nanoparticles . In the FNS theory , cooperativity parameter m may be of particular value . It allows for the direct comparison of energetic cooperativity magnitude differences between related nanoparticles that might be undergoing rationale design by investigators . As such , it could represent an important evaluation tool for structural alterations made with the aim to ultimately achieve optimal mechanical and energetic properties of natural and synthetic nanocompartments . In the case of natural viral nanoparticles , FNS theory may aid in revealing how mechanical properties correlate with local conformational dynamics of the capsid structure to contribute to crucial steps in the viral infectious cycle , such as receptor binding , genome uncoating and capsid maturation .
In our MD simulation studies , we employed multiscale modeling , which combines the simulations of atomic structural models [35] with amino acid residue ( Cα-atom ) based Self Organized Polymer ( SOP ) model of biological particles [24 , 25 , 36 , 37] . In this approach , we first use the all-atom Molecular Dynamics simulations of atomic structural models of a biological particle in question in implicit water using the Solvent Accessible Surface Area ( SASA ) model and Generalized Born ( GB ) model of implicit solvation . These equilibrium MD simulations are carried out in order to obtain an accurate parameterization of the SOP model , as described in the S4 Text . The atomic-level details that determine the type and number of residue-residue contacts between amino acids and their energies are then ported to the SOP model of the particle structure . In dynamic force measurements in silico , the cantilever base is represented by the virtual particle , connected to the spherical bead of radius Rtip , mimicking the cantilever tip ( indenter ) , by a harmonic spring ( S2 Fig ) . The tip interacts with the particles via the repulsive Lennard-Jones potential: U t i p = ∑ i = 1 N ε t i p σ t i p | r i - r t i p | - R t i p 6 ( 19 ) thereby producing an indentation on the particle’s outer surface . In Eq ( 19 ) , ri and rtip are coordinates of the i-th particle and the center of the tip , respectively , εtip = 4 . 18 kJ/mol , and σtip = 1 . 0 Å are parameters of interaction , and the summation is performed over all the particles under the tip . For the cantilever tip ( sphere in S2 Fig ) , we solve numerically the following Langevin equation of motion: η d r t i p d t = - ∂ U t i p ( r t i p ) ∂ r t i p + κ ( ( r t i p 0 - ν f t ) - r t i p ) ( 20 ) where r t i p 0 is the initial position of spherical tip center ( νf is the cantilever base velocity; κ is the cantilever spring constant ) , and the friction coefficient η = 7 . 0 × 106 pN ps/nm . To generate the dynamics of the biological particle of interest tested mechanically , we solve numerically Eqs . ( S1 ) — ( S5 ) for the particle ( see S1 Text ) and Eqs ( 19 ) and ( 20 ) for the indenter ( spherical tip ) . The cantilever base moving with constant velocity ( νf ) ( S2 Fig , S1 Movie ) exerts ( through the tip ) the time-dependent force ( force-ramp ) f ( t ) = f ( t ) n in the direction n perpendicular to the particle outer surface . The force magnitude , f ( t ) = rf t , exerted on the particle increases linearly in time t with the force-loading rate rf = κνf . In the simulations of “forward indentation” , the cantilever base ( and spherical tip ) is moving towards the virus capsid . We control the piezo ( cantilever base ) displacement Z , and the cantilever tip position X , which defines the indentation depth ( deformation ) . The resisting force of deformation F from the virus particle , which corresponds to the experimentally measured indentation force is calculated using the energy output from simulations . To prevent the capsid from rolling , we constrain the bottom portion of the particle by fixing selected Cα-atoms contacting the substrate surface . The experimental FZ-spectra were obtained as described in our previous studies [8 , 12 , 38] . In short , hydrophobic glass slides were treated with an alkylsilane [2] . The viral samples were kept under liquid conditions at all times; all the experiments were performed at room temperature . Capsid solutions were incubated for ∼30 minutes on the hydrophobic glass slides prior to the indentation experiments . Olympus OMCL-RC800PSA rectangular , silicon nitride cantilevers ( nominal tip radius <20 nm and spring constant of 0 . 05 N/m ) were calibrated in air yielding a spring constant of κ = 0 . 0524±0 . 002 N/m . Viral imaging [39] and nanoindentation [1] were performed on a Nanotec Electronica AFM ( Tres Cantos , Spain ) . For empty CCMV , νf = 0 . 06 μm/s , Rtip = 20 nm , and κ = 0 . 05 N/m . For empty TrV , νf = 0 . 06 μm/s , Rtip = 15 nm , and κ = 0 . 056 N/m . For TrV with ssRNA , νf = 0 . 06 μm/s , Rtip = 15 nm , and κ = 0 . 1 N/m . For full AdV with dsDNA , νf = 0 . 055 μm/s , Rtip = 15 nm , and κ = 0 . 0524 N/m . The indentation data were analyzed using a home-written Labview program ( National Instruments ) as described elsewhere [38] . To obtain force-deformation spectra ( FX-curves ) from the experimental output ( FZ-curves ) , we employed the coordinate transformation from the Z-representation ( FZ-curves ) to the X-representation ( FX-curves ) , i . e . X = Z−F/κ [40] .
|
Dynamic force experiments , which have become available to explore the physical properties of biological assemblies , oftentimes reveal results that are difficult to understand without theoretical framework . We employed a multiscale modeling approach—a combination of Molecular Dynamics simulations of atomic structures with Langevin simulations of coarse-grained models of virus shells—to characterize the degrees of freedom defining the deformation and structural collapse of biological particles tested mechanically . This enabled us to develop an analytical model that provides meaningful interpretation of force-deformation spectra available from single-particle nanoindentation experiments . The Fluctuating Nonlinear Spring ( FNS ) model of uniaxial particle’s deformation captures essential features of the force-deformation spectra as observed in nanomanipulations in vitro and in silico: initial non-linearity , then a subsequent force decrease transition due to structural collapse . Our theory uniquely combines the elements of continuum mechanics with the statistics of extremes , enabling one to gather mechanical and statistical characteristics of nanoparticles , which determine the Hertzian deformation of the particle’s protein layer , and bending deformation and structural damage to the particle structure . We have demonstrated how the FNS theory can accurately model the deformation of several viral shells , showing promising model applications for describing a variety of natural and synthetic nanoparticles .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods",
"and",
"Models"
] |
[
"classical",
"mechanics",
"engineering",
"and",
"technology",
"symmetry",
"microbiology",
"viral",
"structure",
"geometry",
"simulation",
"and",
"modeling",
"mathematics",
"nanoparticles",
"damage",
"mechanics",
"nanotechnology",
"research",
"and",
"analysis",
"methods",
"bending",
"deformation",
"biophysics",
"physics",
"virology",
"biology",
"and",
"life",
"sciences",
"capsids",
"physical",
"sciences",
"computational",
"biology",
"biophysical",
"simulations"
] |
2016
|
Fluctuating Nonlinear Spring Model of Mechanical Deformation of Biological Particles
|
The cytomegalovirus resistance locus Cmv3 has been linked to an epistatic interaction between two loci: a Natural Killer ( NK ) cell receptor gene and the major histocompatibility complex class I ( MHC-I ) locus . To demonstrate the interaction between Cmv3 and H2k , we generated double congenic mice between MA/My and BALB . K mice and an F2 cross between FVB/N ( H-2q ) and BALB . K ( H2k ) mice , two strains susceptible to mouse cytomegalovirus ( MCMV ) . Only mice expressing H2k in conjunction with Cmv3MA/My or Cmv3FVB were resistant to MCMV infection . Subsequently , an F3 cross was carried out between transgenic FVB/H2-Dk and MHC-I deficient mice in which only the progeny expressing Cmv3FVB and a single H2-Dk class-I molecule completely controlled MCMV viral loads . This phenotype was shown to be NK cell–dependent and associated with subsequent NK cell proliferation . Finally , we demonstrated that a number of H2q alleles influence the expression level of H2q molecules , but not intrinsic functional properties of NK cells; viral loads , however , were quantitatively proportional to the number of H2q alleles . Our results support a model in which H-2q molecules convey Ly49-dependent inhibitory signals that interfere with the action of H2-Dk on NK cell activation against MCMV infection . Thus , the integration of activating and inhibitory signals emanating from various MHC-I/NK cell receptor interactions regulates NK cell–mediated control of viral load .
Natural killer ( NK ) cells play an important role in the innate immune response against tumors , MHC-mismatched bone-marrow grafts , and pathogens [1]–[2] . These cells also contribute to defense against parasites and intracellular bacteria , and they are critical for the control of a variety of viral infections [3]–[6] . NK cell actions are immediate and appear to be particularly important during the first few days of infection; they involve direct lysis of infected cells and production of proinflammatory cytokines [7] . NK cell activation is tightly regulated by output signals derived from the engagement of inhibitory and activating receptors by their respective ligands on potential targets [8] . Inhibitory human killer immunoglobulin–like receptors ( KIR ) , mouse killer C-type lectin-like receptors family A ( KLRA or Ly49 ) , and NKG2A/CD94 receptors recognize major histocompatibility ( MHC ) class I molecules ( H2 in mice ) , thus controlling NK cell reactivity against “self . ” As virally infected cells downregulate the expression of MHC class I molecules , the lack of inhibitory signals stimulates NK cells . This mechanism is described as the “missing self” hypothesis , whereby NK cells eliminate targets that lack normal levels of self-MHC class I molecules [9] . In addition , the interaction between inhibitory receptors and self MHC-I molecules is the basis of NK cell education ( also termed licensing ) , leading to the maturation of functional NK cells in homeostatic conditions [10]–[17] . By contrast , several families of activating receptors , such as activating KLRA ( also known as Ly49 ) receptors , KLRK1 ( NKG2D ) and the natural cytotoxicity receptor ( NCR ) NKP46 ( NCR1 ) can induce NK cell activation through the recognition of viral ligands or stress-induced molecules [18]–[22] . Although it is clear that NK cell responses are modulated by a balance of opposing signals received from self- or nonself-specific ligands , the precise contribution of specific inhibitory and activating pathways to the resolution of infection remains to be fully understood . The genetic dissection of host resistance or susceptibility to mouse cytomegalovirus ( MCMV ) has provided a fresh view of the precise role of activating NK cell receptors in the recognition of infected cells and host protection against the infection . Using informative crosses between various mouse strain combinations , several MCMV-resistance loci have been mapped to the NK cell gene complex ( NKC ) on mouse chromosome 6 . The best characterized , Cmv1 ( also known as Klra8 ) and Cmv3 , are defined by two different modes of inheritance , which seem to correlate with two different mechanisms of recognition . Cmv1 is a single dominant locus whose resistance allele , described in C57BL/6 ( B6 ) mice , encodes the Ly49H activating receptor . Ly49H recognizes MCMV-infected cells through a direct interaction with the viral product m157 [21]–[22] . Engagement of Ly49H by m157 elicits NK cell–mediated cytotoxicity , cytokine secretion , NK cell proliferation , and viral clearance [18] , [23]–[24] . The Cmv3 locus was detected in a cross between resistant MA/My and susceptible BALB/c mice . Expression of Cmv3-determined resistance accounted for a 100-fold decrease in splenic viral load , but it was only observed in mice carrying a specific combination of MA/My alleles at the NKC and MHC ( H2k ) loci . Functional candidate gene testing of Ly49 receptors isolated from MA/My mice showed that another DAP12-associated receptor , Ly49P , responded to MCMV-infected cells [25] . In this case , Ly49P functional recognition of target cells required surface expression of both the host H2-Dk molecule and the viral component m04/gp34 [26] . The role of the H2k haplotype in MCMV resistance was previously associated with greater survival following infection with lethal inoculum doses of MCMV compared to other H2 haplotypes [27] . In addition , linkage analyses in a cross between resistant MA/My and susceptible C57L strains , as well as the generation of congenic C57L . M-H2k mice carrying the H2k allele from MA/My , confirmed a role for H2-Dk-linked resistance to MCMV [28]–[29] . Nevertheless , the mechanism of resistance regulated by the interaction between NK receptors and MHC class I molecule is still unclear . In MA/My mice , Cmv3-determined MCMV resistance served as a model for researchers and allowed them to propose the existence of a functional interaction between the activating Ly49P receptor on NK cells and MHC class I H2-Dk molecules . However , the role of a Ly49P-m04-H2-Dk stimulatory axis remained to be clarified . In the present study , we sought to replicate experimentally the statistical association between the NKC and the MHC-I locus in MCMV resistance , to determine the precise molecules involved in MCMV resistance in vivo , and to evaluate the impact of MHC-I inhibitory signals on the NK cell antiviral response .
To validate the epistatic interaction between the NKC and H2 detected by linkage analysis [25] , we used a marker-assisted strategy to construct congenic mouse lines in which a chromosome 6 segment ( Cmv3 ) from MA/My MCMV-resistant mice was independently introgressed into BALB/c ( H2d ) and BALB . K ( H2k ) susceptible backgrounds . Congenic BALB . K mice have been described previously [30] . The correlations between the current physical maps and the genomic region introgressed in the respective single and double congenic strains BALB-Cmv3MA/MyH2d and BALB-Cmv3MA/MyH2k are shown in Figure 1 and Table 1 . To examine the effect of the genetic background on the expression of NKC-encoded receptors in the Cmv3MA/My region , we used a panel of antibodies with known antigen specificities ( Figure S1 ) [31] . Though several anti-Ly49 antibodies are cross-reactive [31] , we observed variations among the mouse strains in terms of frequency of Ly49 subpopulations . Compared to BALB-Cmv3MA/MyH2d mice , BALB-Cmv3MA/MyH2k animals had a significantly increased frequency of NK cells stained with the monoclonal antibodies 12A8 ( against Ly49R; P = 0 . 02 ) and 14B11 ( against Ly49I/U; P = 0 . 007 ) ( Figure 1B and Figure S2 ) . Notably , 14B11-stained NK cells were also significantly increased in BALB-Cmv3MA/MyH2k mice compared to MA/My mice ( P = 0 . 005 ) . These results demonstrated the influence of H2 alleles [32] , as well as the influence of an additional non-H2 mechanism , in the formation of the Ly49 repertoire . We also observed a highly significant increase in the frequency of NK cells labeled with 2F1 antibody ( P = 0 . 004 ) , which recognizes the maturation marker KLRG1 , in BALB-Cmv3MA/MyH2d mice compared to their H2k counterparts . Finally , we observed that NK cells from the three mouse strains that share the Cmv3MA/My allele lacked expression of NKG2A/C/E and CD94-associated receptors at the protein but not mRNA level ( Figure 1B and Figure S3B ) . By contrast , there was a normal expression of these receptors in the FVB/N mouse strain , which carries an NKC haplotype similar to that of MA/My mice ( Figure S3A and S3B ) . To evaluate the effect of the transferred MA/My chromosome 6 ( Cmv3 ) segment on the response of MCMV-susceptible BALB/c and BALB . K mice , we infected congenic and parental control mice by intraperitoneal ( i . p . ) inoculation of MCMV sublethal doses ( Figure 1C ) . By day 3 post-infection ( p . i ) , uncontrolled MCMV replication was observed in the spleen of susceptible BALB/c mice ( log10 plaque-forming units [PFU] = 4 . 39±0 . 16 ) , while MCMV-resistant MA/My mice had restricted viral replication , as shown by a >100-fold lower viral titer ( log10 PFU = 1 . 81±0 . 05 ) than that seen in BALB/c mice . Single congenic BALB . K and BALB-Cmv3MA/MyH2d mice had viral titers that were indistinguishable from those observed in BALB/c mice . More importantly , congenic mice with the BALB . K background , which carry only one copy of the MA/My NKC ( BALB-NKCBALBCmv3MA/My . H2k ) , were as susceptible to MCMV as BALB/c , BALB . K , and BALB-Cmv3MA/MyH2d mice . By contrast , double congenic BALB-Cmv3MA/MyH2k mice had restricted virus growth to the same extent as resistant MA/My mice . Viral titers in the liver correlated with those observed in the spleens . Furthermore , by day 7 p . i . , the virus was cleared from the spleen ( unpublished data ) , which had undergone a massive increase in weight and cell number in MA/My and BALB-Cmv3MA/MyH2k mice ( Figure 1D ) . Collectively , these data demonstrate that the interaction between Cmv3MA/My and H2k confers MCMV resistance and is sufficient to explain the control of viral load observed in MA/My mice . Because we did not have antibodies that specifically recognize Ly49P receptors and to examine a possible role of CD94 heterodimers , we attempted to confirm the results obtained in the congenic mice in a new cross between two strains that independently carried the H2k loci and the Ly49P gene at the NKC . We examined the segregation of MCMV viral load in the spleens of progeny mice from an F2 cross between the MCMV-susceptible strains FVB/N and BALB . K . Although both parental strains sustained a relatively high viral titer ( 5 . 2 log10 PFU ) , the 137 F2 progeny showed a continuous distribution ranging from 2 to 6 log10 PFU ( Figure 2A ) . To evaluate the contribution of H2 and NKC genes to MCMV resistance in this cross , F2 mice were genotyped and distributed according to their NKC ( Ly49e ) and H2 ( IAA1 ) genotypes . Mice homozygous for H2kk alleles from BALB . K and NKCff alleles from FVB/N had the lowest viral load ( Figure 2B ) . The model that best fitted this phenotype/genotype distribution in the analysis of variance had a joint logarithm of odds ( LOD ) score of 9 ( P<10−11 ) and accounted for 29 . 6% of the phenotypic variation ( Table 2 ) . Thus , there was a highly significant association between NKC/H2 interaction and control of MCMV infection in this second cross , indicating that Cmv3 was also present in the FVB/N mouse strain and that its expression in the presence of H2k was necessary and sufficient for viral control . Furthermore , these data suggest that the same gene encodes Cmv3 in both the MA/My and FVB/N NKC regions . Previously , we showed that activation of Ly49P-bearing reporter cells requires the H2-Dk host molecule on MCMV-infected cells [25]–[26] . Therefore , to better delineate the role of H2 in the host response against MCMV , we attempted in vivo rescue of the FVB/N susceptible phenotype by genetic transfer of an 11 kb H2-Dk genomic fragment cloned from AKR mice ( Figure 3A ) . We monitored for the presence of a diagnostic 300 bp fragment corresponding to exon 3 of the H2-Dk gene to identify transgenic FVB-Tg ( Dk ) + mice among the founder population ( unpublished data ) . By surface staining of mouse embryo fibroblasts ( MEF ) from FVB-Tg ( Dk ) + mice , we observed the normal low levels of H2-Dk expression under regular conditions ( Figure 3B ) . However , IFN-β treatment up-regulated expression of H2-Dk on MEF cells from either FVB-Tg ( Dk ) + mice or AKR mice ( H2-Dk transgene donor mouse strain ) to the same extent , indicating that the transgene promoter regulatory sequences were intact ( Figure 3B ) . We also found that the level of expression of H2-Dk in splenocytes from the FVB/N transgenic mice was similar to the natural H2-Dk expression in splenocytes from MA/My or BALB . K strains ( Figure 3C ) . Finally , we investigated the expression of transgenic H2-Dk and endogenous H2-Dq molecules in T and B cells isolated from the spleen and observed that the two MHC-I molecules were expressed in FVB-Tg ( Dk ) + mice at levels similar to those found in H2q- and H2k-bearing inbred mice ( Figure S4 ) . To test whether the H2-Dk transgene is capable of stimulating the Ly49P receptor , we assessed the activation of Ly49P-bearing reporter cells and found that these cells were equally activated by MCMV-infected MEF cells from BALB . K or from FVB-Tg ( Dk ) + ( Figure S5A ) . Stimulation of Ly49P reporters was also observed upon challenge by MEFs infected with a mutant virus lacking m157 ( the Ly49H ligand ) . However , Ly49P reporter cell stimulation was lost upon infection of MEF cells with a mutant virus lacking the m04 gene , indicating that the transgenic H2-Dk molecule also requires m04/gp34 to stimulate Ly49P , as reported ( Figure S5B ) [26] . To establish the contribution of H2-Dk to the MCMV response , we first investigated a possible modulation of the Ly49 receptor repertoire . No significant differences were found between transgenic and non-transgenic mice in terms of the frequency of the various NK cell populations tested ( Figure S6 ) . Since H2-Dk has the potential to influence licensing through its interactions with cognate Ly49 inhibitory receptors [16] , we addressed the licensing status of wild-type and transgenic NK cells . To do this , we determined the ability of NK cells to mediate in vivo rejection of MHC class I deficient splenocytes isolated from B6 . 129P-H2-D1tm1Bpe H2-K1tm1Bpe ( herein B6 . H20 ) mice using a quantitative CFSE-based method [33] . As previously described , B6 . H20 mice tolerated grafted syngeneic splenocytes ( Figure 3D ) . By contrast , both FVB-Tg ( Dk ) + and FVB-Tg ( Dk ) − mice rejected B6 . H20 splenocytes with the same efficiency , suggesting that the presence of the H2-Dk transgene does not alter the licensing status of NK cells ( Figure 3D ) . Finally , we monitored early viral replication following MCMV infection in FVB-Tg ( Dk ) + mice , along with single and double BALB-Cmv3MA/MyH2d and BALB-Cmv3MA/MyH2k congenic mice and parental MA/My mice . We observed that FVB-Tg ( Dk ) + mice had a statistically significant , albeit modest , reduction in MCMV replication in the spleen ( but not in the liver ) compared to nontrangenic littermates; this reduction represented only a small fraction ( 1∶7 ) of the reduction observed in BALB-Cmv3MA/MyH2k congenic mice ( Figure 3E ) . Collectively , these data demonstrated that the H2-Dk transgene was fully expressed and able to recognize and activate the Ly49P receptor in vitro; however , it only provided partial control of MCMV infection . Classical MHC class I molecules are the prototype ligands for Ly49 receptors . Reporter cell assays and tetramer binding assays suggest that H2-Dk molecules elicit both activating signals , through Ly49P , and inhibitory signals , through Ly49I and Ly49V [25] , [31] . By contrast , H2q-encoded molecules can elicit strong inhibitory signals through Ly49I or Ly49C , but are inert to Ly49G2 [34] , as well as to Cmv3-encoded activating receptors ( Ly49P , Ly49R , and Ly49U ) [25] . Therefore , we hypothesized that the poor MCMV infection control observed in FVB-Tg ( Dk ) + mice resulted from competition between the inhibitory and activating signals emanating from H2q- and H2-Dk-encoded ligands , respectively . To test this hypothesis , we crossed FVB-Tg ( Dk ) + transgenic mice with B6 . H20 mice , which possess targeted deletions at the H2-D and H2-K genes . This cross produced F3 progeny mice homozygous for the FVB/N NKC ( Cmv3FVB ) , but with a different assortment of MHC class I alleles . F3 mice were either ( 1 ) deficient in endogenous MHC class I alleles in the presence or absence of the H2-Dk transgene ( H20/Tg ( Dk ) + or H20/Tg ( Dk ) − ) , ( 2 ) hemizygous for H2q in the presence or absence of the H2-Dk transgene ( H20/q/Tg ( Dk ) − or H20/q/Tg ( Dk ) + ) , or ( 3 ) homozygous for H2q in the presence of the H2-Dk transgene ( H2q/Tg ( Dk ) + ) ( Figure 4A and Table 1 ) . Again , we monitored whether the frequencies of various NK cell populations were affected by the genetic makeup of F3 mice and detected no major variations in the NK cell populations , the only exception being Ly49G+ NK cells , which were barely detectable in Cmv3FVB/H20/Tg ( Dk ) − mice ( Figure S7 ) . Similarly , the level of H2-Dk and H2-Kk expression on splenocytes was equivalent in transgenic F3 mice with different H2 genotypes ( Figure S7 , right panel ) . By contrast , we noted that levels of H2-Dq expression on lymphocytes from homozygous H2q/Tg ( Dk ) + transgenic mice were double those of hemizygous H20/q/Tg ( Dk ) + transgenic mice ( P = 0 . 001 ) ( Figure 4B and bar graph ) . Despite this variation in MHC class I expression levels , licensing of NK cells from H2-Dk transgenic mice carrying no ( 0 ) , one ( 0/q ) , or two ( q/q ) H2q alleles was equivalent , as shown by their ability to reject CFSE-labeled splenocytes from B6 . H20 mice [Figure 4C right] . These results were confirmed using explanted , IL-2-activated NK cells in cytotoxicity assays against MCH class I–deficient RMA/S target cells [Figure 4C left] , demonstrating that NK cells from transgenic mice with different H2 assortments sense equally the loss of MHC class I expression on target cells and therefore are equally educated [16] . To determine the influence of the various MHC class I molecules on the NK cell immune response to MCMV infection , we examined F3 mice and parental controls at early time points , particularly on day 3 p . i . , when receptor-specific NK cell responses are established [18] . On day 3 , the post-infection viral titers in the spleens and livers of Cmv3FVB/H20/q/Tg ( Dk ) + and Cmv3FVB/H20/Tg ( Dk ) − mice were indistinguishable , demonstrating that the presence of H2q dampens the effect of Cmv3FVB/H2k on the containment of virus replication ( Figure 4D ) . By contrast , the presence of the transgene had a significant effect in the absence of endogenous class I molecules , as Cmv3FVB/H20/Tg ( Dk ) + mice had close to 30-fold lower viral titers compared to Cmv3FVB/H20/q/Tg ( Dk ) − mice . In parallel , B6 . H20 control mice , which express Ly49H , also cleared the virus load despite lacking MHC-I molecules ( Figure 4D ) . To investigate the role of NK cells in limiting viral spread , we found that , as in MA/My mice , the control of virus load was abrogated in BALB-Cmv3MA/MyH2k and Cmv3FVB/H20/Tg ( Dk ) + mice if treated with anti-asialo GM1 antibody prior to MCMV infection , demonstrating that the resistance phenotype is NK cell-dependent ( Figure S8A ) . Indeed , we observed uncontrolled virus growth not only when MA/My mice were pretreated with anti-asialo GM1 and anti-NK1 . 1 antibodies , but also after they were pretreated with YE1/48 ( anti-Ly49PRTV ) , 12A8 ( anti-Ly49R ) , or 4D11 ( anti-Ly49GT ) antibodies , indicating an overlap in Ly49 receptor expression on NK cells ( Figure S9 ) [25] . At day 6 p . i . , a time characterized by robust proliferation of receptor-specific NK cell populations responding to the virus [18] , we found that mice expressing Cmv3 resistance ( MA/My , BALB-Cmv3MA/MyH2k , and Cmv3FVB/H20/Tg ( Dk ) + ) had cleared MCMV from the spleen ( unpublished data ) . Furthermore , spleen cell numbers were increased 2–6-fold in these mice and BrdU uptake indicated a robust NK cell proliferation ( Figure S8B ) . Together , our results indicate that expression of H2-Dk can rescue Cmv3-determined MCMV resistance in the absence of endogenous H2q molecules and that Cmv3/H2-Dk-mediated resistance is associated with the expansion of NK cells in response to infection . To better define the role of H2q alleles , we compared the kinetics of viral replication in Cmv3FVB-Tg ( Dk ) + transgenic mice carrying no , one , or two H2q alleles . We observed that the number of H2 alleles correlated with a quantitative increase in viral load , as early as 36 hours p . i . . On days 3 and 5 p . i . , differences in viral containment among mice of the three genotypes were significant ( Figure 5A ) . To investigate the effect of H2q molecules on NK cell specific responses against MCMV , we monitored BrdU incorporation on NK cells after MCMV infection in FVB/N WT and F3 mice carrying no , one , or two copies of H2q alleles . After 5 days p . i . , NK cells were stained with the anti-Ly49ORV ( 4E5 ) monoclonal antibody , which stained around 50% of NK cells in these strains ( Figure S6 and Figure S7 ) , and with the anti-BrdU monoclonal antibody . In all mice , we observed that NK cells that incorporated BrdU were negative for the Ly49ORV antibody staining . Furthermore , the increase in BrdU incorporation was inversely proportional to the number of H2q alleles ( Figure 5B ) . This result suggests that there is a dose-dependent inhibition of NK cell proliferation by H2q alleles in response to MCMV infection . Finally , to investigate whether host MHC-I molecules affect NK cell activity upon MCMV infection , we adoptively transferred CFSE-labeled NK cells enriched from Cmv3FVB/H20/Tg ( Dk ) + donor mice into Cmv3FVB/H20/Tg ( Dk ) + or Cmv3FVB/H20/q/Tg ( Dk ) + recipients ( Figure 5C ) . After 5 days p . i . , donor NK cells had undergone more rounds of division in the Cmv3FVB/H20/Tg ( Dk ) + recipients than the NK cells that were transferred into Cmv3FVB/H20/q/Tg ( Dk ) + mice; this indicated that H2q alleles limited NK cell proliferation induced by MCMV infection ( Figure 6C bar graph ) . Thus , NK cells carrying H2-Dk as the sole MHC class I molecule were impaired in their ability to proliferate if the recipient mice carried H2q alleles . Collectively , these results suggest that expression of host H2q molecules dampens the capacity of NK cells to protect against MCMV .
The role of the MHC has been studied using panels of congenic [27] , sub-congenic , and transgenic mice or F2 crosses with the same NKC haplotype [28]–[29] , [35] . To the best of our knowledge , the present study is the first to provide formal proof of the impact of both NKC and MHC haplotypes on NK cell antiviral activities in vivo . Our study , through the use of single- and double-congenic mice , minimized differences in non-NKC or non-MHC genes . Thus , we established that the joint action of specific alleles at these two regions accounted for most of the overall phenotypic differences between the MCMV-susceptible BALB/c and MCMV-resistant MA/My mouse strains . BALB-Cmv3MA/MyH2k mice were indistinguishable from MA/My mice in terms of initial control of infection and late NK cell responses . Although the MA/My NKC region in congenic mice encompasses more than just Ly49 genes , our data indicates that the influence of MHC alleles on MCMV-resistance stems from the capacity of MHC class I molecules to serve as ligands for Ly49 receptors . In support of this hypothesis , the F2 cross between the MCMV-susceptible FVB/N ( H2q ) and BALB . K ( H2k ) mouse strains demonstrated that FVB/N mice carried a Cmv3 resistance allele that was conditional to H2k and overridden by the H2q susceptibility allele . Within the Cmv3 region , Ly49 receptors were responsive to MHC class I ligands . On the one hand , we noticed that none of the available anti-NKG2 or anti-CD94 antibodies recognized MA/My mouse NK cells , in contrast to NK cell recognition in FVB/N mice . Nevertheless , we observed that F2 mice of the combined Cmv3FVB/H2k genotype restrained viral replication to an extent similar to that seen in MA/My mice . On the other hand , our haplotype studies [25] , [36] and new public data ( http://phenome . jax . org ) indicate that FVB/N and MA/My share the same Ly49 gene repertoire , including Ly49P . Consequently , it seems that NK cell responsiveness during MCMV infection varies with different NKC-MHC combinations and is optimal only with a precise combination of Ly49 receptors inherited from MA/My ( or FVB/N ) mice and MHC class I H2k molecules . Our results confirmed that the H2 effect was due to the MHC class 1 molecule H2-Dk . Using an 11 kb genomic fragment containing a functional H2-Dk gene , we achieved a phenotypic rescue , although the rescue was incomplete if combined with H2q alleles . The complete protective effect of H2-Dk was restored in F3 mice lacking endogenous H2q molecules . Although H2-Dk also affects the adaptive immune response , early containment of viral replication , massive NK cell proliferation , and reversal of the resistance phenotype by depletion of NK cells in FVB-H20-Tg ( Dk ) + clearly support a mechanism at the level of NK cells . Because of the presence of both inhibitory and activating Ly49 receptors , several nonexclusive scenarios could account for the precise mode of action of the combined MHC class I H2-Dk and Ly49 genotypes on the NK cell response against MCMV: ( 1 ) low threshold of NK cell activation through weak H2-Dk/Ly49 inhibitory signals , ( 2 ) effective NK cell activation through H2-Dk/Ly49 activating signals , and ( 3 ) interplay between H2-Dk/Ly49 activating and inhibitory signals . One possibility is that MHC class I/inhibitory Ly49 signals have a negative impact on the NK cell response to MCMV . In our study , mature NK cells in BALB . K mice ( which are the most susceptible to MCMV infection ) express three inhibitory receptors: Ly49A , Ly49C , and Ly49G2 , which all bind to MHC-I H2k molecules [31] , [34] , [37] . Thus , the majority of NK cells from BALB . K mice should be inhibited by a receptor for a self-ligand . Indeed , we have recently shown that deletion of the m04 gene renders BALB . K mice resistant to MCMV infection , as the protein it encodes abolishes NK cell activation via the “missing-self” recognition mechanism ( Babic et al . , 2010 ) [56] . The m04/gp34 protein escorts MHC class I molecules to the surface of infected cells , thus maintaining a level of surface MHC expression sufficient enough to trigger inhibitory NK cell receptors [38] . Thus , with only three ( Ly49V , Ly49I , and Ly49G2 ) out of seven Ly49 inhibitory receptors able to recognize H2-Dk molecules , NK cells from BALB . Cmv3MA/My H2k mice should be less susceptible to inhibition by H2k binding ( Figure 6A–6C ) . The existence of an H2-Dk–mediated activating axis to MCMV resistance is supported by the gain-of-function phenotype of FVB-H2-Dk transgenic mice , which presented itself despite their Ly49 repertoire that is virtually identical to that of their non-transgenic littermates ( Figure S6A and S6B ) . Furthermore , the absence of NK cell triggering through inhibitory Ly49 receptors was not sufficient to allow efficient control of MCMV replication , as demonstrated by the F3 Cmv3FVB MHC class I–deficient mice . Most NK cells that develop in MHC class I–deficient hosts are unable to respond to MHC class I–deficient targets . However , a recent study demonstrated that , in the context of MCMV infection , NK cells eliminate virally infected cells in MHC class I–negative hosts , in addition to regaining the ability to eliminate MHC class I–deficient hematopoietic host cells [39] . This mechanism seems to be triggered by the inflammatory milieu induced by MCMV infection [39] . These observations suggest that the susceptibility of Cmv3FVB MHC class I–deficient F3 animals to MCMV infection is not due to a defect in education but to the absence of an activation axis , which is provided by H2-Dk . Activating signals , mediated by the engagement of Ly49P by H2-Dk/m04 , provided only a marginal enhancement of the NK cell response in the presence of H2q . Interestingly , we observed a gene dosage effect in the inhibitory action of H2q that correlated with the level of surface expression of this MHC class I molecule . However , H2q copy number did not affect the ability of NK cells from H2-Dk transgenic mice ( FVB or F3 ) to eliminate MHC class I deficient target cells; this indicates that H2q gene dosage does not alter education/licensing of NK cells . By contrast , adoptive transfer experiments demonstrated that H2q alleles expressed on host cells limit the ability of NK cells to respond to MCMV infection , indicating that the H2q effect influences NK cell recognition of class I ligands on target cells . This suggests that H2q inhibitory signals dominate over H2-Dk-dependent activating signals emanating from MCMV-infected cells . One possibility is that H2q inhibitory signals are stronger and/or more frequent than H2k-dependent activating signals . Indeed , it has been shown that both the density and the avidity of inhibitory Ly49-ligand pairs determine the strength of inhibition [40] . Alternatively , H2q MHC class I molecules could compete with H2-Dk for binding with the m04 protein and thus blunt the m04/H2-Dk-Ly49P activating axis . We have noted that Ly49P reporter cells are equally stimulated by MCMV-infected MEFs of H2k or H2k/q genotype , which may indicate otherwise ( Figure S5 ) . However , these results might not reflect the effect of H2q molecules on the H2-Dk/m04 complex under physiological conditions . While the molecular details of H2q inhibition of NK cell function remain unclear , our results suggest a model in which two antagonistic mechanisms are at play in NKC-H2-determined resistance to virus infection ( Figure 6 ) . One involves enhanced NK cell responses through H2-Dk-mediated activating signals . The other involves dampened NK cell responses through inhibitory Ly49 receptors stimulated by class I H2q ( or H2d ) molecules , which override the effect of the H2-Dk construct . It is puzzling that Ly49 receptors can sense MHC class I molecules on infected cells despite immune-evasion mechanisms elaborated by MCMV that downregulate surface expression of MHC class I molecules . Indeed , mouse strain–specific [41] and cell type–specific [42] differences have been reported in the ability of immunoevasins to inhibit lysis of infected cells by CTLs , indicating that the efficiency of MHC class I downregulation during MCMV infection [43] is context dependent . In vivo MCMV replication occurs in a multitude of cell types , and perhaps the ability of the virus to achieve immune avoidance selectively might contribute to the delicate equilibrium of coexistence it has established with the host . The striking similarities between Ly49 and KIR interactions with their respective MHC-I ligands and how they both affect NK cell function prompted us and other researchers to use the mouse as a model to study NK cell antiviral responses . Our results lend support to clinical and epidemiological studies implicating KIR-HLA interactions of different strengths in determining a hierarchy of NK cell activation with varied effects on the host response against herpersviruses [44] , HCV [45] , and HIV [46] . Our work also highlights the ability of inhibitory signals to overcome NK cell activation . These regulatory mechanisms would be relevant in conditions where NK cell activation is undesirable during infection or immune disease . For example , activating KIR genotypes have been found to predispose to reactivation of quiescent , opportunistic infections associated with herpesvirus infections in HIV patients [47] ) , and to fatal outcome following Ebola virus infection [48]; furthermore , they may constitute a risk factor for susceptibility to autoimmunity and certain cancers [49] , [50] . Ultimately , our data indicate that , as has been proposed for cancer and autoimmunity , manipulating the balance between inhibitory and activating NK receptor signals represents a possible avenue to harness the therapeutic potential of NK cells against virus infections .
The animal protocols and experiments were approved by the Canadian Council on Animal Care ( CCAC ) and the McGill University Animal Resources Center . MA/My , BALB . K , BALB/c , C57Bl/6 ( B6 ) , DBA/J , and AKR mice were purchased from The Jackson Laboratory . FVB/N mice were purchased from Charles Rivers Laboratories . The B6 mice deficient for H2-DbKb ( B6 . H20 ) were kindly provided by Hidde L . Ploegh ( Cambridge , Massachusetts ) . BALB-Cmv3MA/MyH2k and BALB-Cmv3MA/MyH2d were generated by backcrossing the ( MA/MyXBALB . K ) F1 or ( MA/MyXBALB/c ) F1 into BALB . K or BALB/c , respectively , for at least ten generations . At each backcross , inheritance of the NKC from parental MA/My mice was genotyped using either the Ly49e marker or the D6mit135 marker [51] . In the progeny , the introgressed portion from parental MA/My mice , which included the NKC , was analyzed using microsatellite markers or by detecting known SNPs . Once the genetic region was reduced from 34 Mb ( between D6MIT36 and D6MIT59 ) to 10 Mb ( between rs13479016 and rs13479061 SNPs ) , heterozygous mice were intercrossed to generate the homozygous congenic lines . The H2-Dk genomic fragment cloned into the PBR22 plasmid was kindly provided by Bernd Arnold ( Deutsches Krebsforschungszentrum [DKFZ] , Heidelberg , Germany ) . The 11 . 5 kb fragment encompassing the Dk gene was subsequently purified and injected into fertilized FVB/N mouse eggs . Transgenesis was performed at the Quebec Transgenesis Research Network ( QTRN ) . Transgenic founders were screened by PCR with the primers 5′-cacacgatccagcggctgt-3′ and 5′-ggcccggtctctctctgcag-3′ , specific for H2-Dk exon 3 . They were then bred to FVB/N WT mice . To generate F3 mice , FVB-Tg ( Dk ) and B6 . H20 mice were bred to produce F1 and F2 progeny . To discriminate between the NKC and H2 regions inherited from the parental strains , the F2 mice were genotyped at the NKC region with the D6Mit61 and D6Mit52 markers and at the H2 region with the D17Mit51 marker; they were also genotyped for the presence or absence of the H2-Dk transgene . Only the mice homozygous for the FVB/N NKC and heterozygous for either the H2 or the H2-Dk transgene were kept to generate the F3 progeny , as listed in Table 1 . To prepare splenic leukocytes , spleens were removed aseptically then gently mashed through a 70 µm nylon mesh ( BD Bioscience ) . Red blood cells were lysed with ammonium chloride ( Sigma ) . To isolate lymphocytes from mouse blood , mice were bled from the cheek; blood was collected in RPMI medium containing 15 mM EDTA . Lymphocytes were collected after gradient centrifugation using Histopaque-1077 ( Sigma ) . Fc receptors were blocked with 2 . 4G2 antibody prior to staining with specific monoclonal antibodies . NK cells were incubated with NKp46 ( conjugated to phycoerythrin [PE] or fluorescein isothiocyanate [FITC] ) and specific monoclonal antibodies against Ly49A-Biot ( YE148 ) , Ly49A/D ( 12A8 ) , Ly49CIH ( 14B11 ) , Ly49D ( 4E5 ) , Ly49G2 ( 4D11 or AT8 ) , NKG2A/C/E ( 20d5 ) or NKG2A/B6 ( 16A11 ) , CD94 ( 18D3 ) , or KLRG1 ( 2F1 ) . NK cells were also incubated with the following isotype control monoclonal antibodies: PE-conjugated golden syrian hamster IgG , FITC- or PE-conjugated mouse IgG2a K , or FITC-conjugated rat IgG2a K ( e-Bioscience ) . H2-Dk and -Dq products were detected by anti-H2-Dk antibody ( 15-5-5 ) from BioLegend and anti-H2-Dq antibody ( KH117 ) from e-Bioscience . To detect incorporated BrdU on NK cells , mice were scarified 5 or 6 days after MCMV infection; cells were first stained for surface antigens ( anti-NKp46 and/or anti-Ly49 receptors ) and then fixed , permeabilized , treated with DNase I , and stained with FITC- or allophycocyanin ( APC ) -conjugated anti-BrdU antibody ( clone 3D4; BD Biosciences ) , according to the manufacturer's protocol . Flow cytometry analysis was performed with a FACSCalibur flow cytometer ( BD Biosciences ) and data were analyzed using CellQuest ( BD Biosciences ) or FlowJo ( Tree Star ) . To assess NK cell proliferation in vivo , NK cells from the spleen were first enriched by negative selection ( Miltenyi Biotec ) , then incubated with 5 µM CFSE for 15 minutes , washed , and resuspended in PBS . The purity of the NK cells ( 55%–70% ) was evaluated by FACS using anti-NKP46 antibody; 2 million NK cells were then injected intravenously into recipient mice 24 hours before infection with MCMV . The proliferation index , indicating the number of divisions of CFSE-labelled NK cells , was determined using the FlowJo software . Stock MCMV from mouse salivary glands was prepared by passaging the virus ( Smith strain ATCC VR-1399 , lot 1698918 ) twice in BALB/c mice . The virus was prepared from a homogenate of salivary glands 21 days p . i . . Mice aged between 7 and 9 weeks were infected intraperitoneally with 2 , 000 PFUs of MCMV . The tissue culture-grown viruses [52] Δm157 MCMV , which lacks the m157 open reading frame ( ORF ) , and Δm04 MCMV , which lacks the m04 ORF , have been previously described [23] , [53] and were kindly donated by Ulrich H . Koszinowski ( Max von Pettenkofer Institute , Munich , Germany ) and Stepan Jonjic ( Rijeka University , Rijeka , Croatia ) . Viral titers of the stock virus or mouse organs ( spleen and liver ) were evaluated in vitro by standard plaque assays on a confluent BALB/c MEF monolayer , as previously described [54] . The MEFs used in this work were generated as previously described [52] , except for FVB-Tg ( Dk ) + transgenic and nontransgenic MEF cells , which were generated from individuals embryos from the progeny of FVB-Tg ( Dk ) ×FVB wild-type mouse crosses and then genotyped for the presence of the H2-Dk transgene . 2B4 reporter cells expressing Ly49H , Ly49P , Ly49C , or Ly49I were generated as previously described [22] , [25] , [55] . MEF cultures from AKR , FVB/N H2-Dk transgenic , FVB/N wild-type , and BALB . K mice were infected with MCMV ( Smith strain ) or Δm157 or Δm04 deletion viruses at a multiplicity of infection ( MOI ) of 1 . 0 for 24 hours; they were used to stimulate 2B4 reporter cells as previously described [25] . GFP was detected by flow cytometry and analyzed using the FlowJo software . Splenocytes from B6 . H20 mice were labeled with 0 . 4 mM CFSE ( CFSE low ) in RPMI medium containing 5% FCS; splenocytes from recipient mice were labeled with 4 mM CFSE ( CFSE high ) in RPMI containing 10% FCS . The splenocytes were then incubated at 37°C for 10 minutes before being washed three times in RPMI containing 10% FCS . Cells ( 5×106 ) of each type were mixed , and the mixture ( 200 µl ) was injected intravenously into recipient mice . After 18 hours , spleens were harvested and red blood cells were lysed . The relative percentage of cells in each CFSE population was measured by FACS as previously described [33] . NK cells from the spleen were expanded for 5 days in RPMI medium supplemented with 1 , 000 U/ml human IL-2 ( NCI Preclinical Repository ) . Cells were washed in RPMI and stained with NKp46 antibody to determine the purity of NK cells and to adjust the number of NK cells among strains . RMA/S cells were labeled with 0 . 4 mM CFSE in RPMI medium containing 5% FCS for 15 minutes at 37°C and washed three times . CFSE-RMA/S and NK cells were cocultured at effector/target ratios of 2∶1 , 4∶1 , 8∶1 , 16∶1 , and 32∶1 for 4 hours at 37°C . Specific lysis was determined by the measure of 7-aminoactinomycin D ( 7-AAD ) incorporation ( BD ) in CFSE-RMA/S cells by flow cytometry , as previously described [56] . For the 137 ( FVB/N×BALB/c ) F2 mice , the contribution of the NKC and H2 loci to the segregation of the phenotype was estimated with the linear model phenotype = m+NKC+H2+NKC:H2+e , where NKC and H2 represent factors that depend on the mode of inheritance proposed , m is the common mean value , NKC:H2 is an interaction term , and e is the independent , normally distributed random deviations . For the additive mode of inheritance , the NKC and H2 represent the number of FVB/N/BALB . K alleles at each locus . For the recessive mode of inheritance , the NKC and H2 are indicator variables of the homozygous FVB/N and BALB . K backgrounds , respectively . The four possible additive-recessive combinations of H2-NKC models , with and without an interaction term , were fitted . We assessed the magnitude of the contribution for each term in the model by its P value , obtained by 1 million bootstrapped samples , and partial η2 . Partial η2 was computed as η2 = SSfactor/ ( SSfactor+SSerror ) , where SSfactor is the type 3 associated sum of squares with the factor in the analysis of variance ( ANOVA ) table , and SSerror is the sum of squares corresponding to the residual variation . We carried out statistical and graphical analyses using R software . For other statistical analyses in this work , differences between groups were calculated with two-way ANOVA analysis , followed by Bonferroni after tests . For some of the analyses , unpaired , two-tailed Student's t-tests were conducted . Results with a P value of <0 . 05 were considered to be statistically significant .
|
Effective natural killer ( NK ) cell responses against virally infected cells are regulated by NK cell receptors that specifically recognize target cells . In the current study , we validated the specific interaction taking place between NK cell receptors and MHC class I molecules on the surface of infected cells , resulting in resistance to cytomegalovirus . Genetic dissection of this mechanism of interaction revealed that the NK cell response occurs exclusively through the triggering of the activating Ly49P receptor by the MHC class I H2-Dk molecule . We observed , in this context , that NK cells were incapable of clearing the virus when target cells also expressed MHC class I H2q molecules , which strongly and quantitatively inhibit NK cells . Our findings reveal that the interplay between inhibitory and activating NK cell receptors and their MHC class I ligands generate signals that shape the outcome of infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/gene",
"discovery",
"virology/mechanisms",
"of",
"resistance",
"and",
"susceptibility,",
"including",
"host",
"genetics",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system",
"genetics",
"and",
"genomics/complex",
"traits",
"immunology/immune",
"response",
"immunology/innate",
"immunity",
"virology/animal",
"models",
"of",
"infection",
"immunology/genetics",
"of",
"the",
"immune",
"system",
"infectious",
"diseases/viral",
"infections",
"virology/host",
"antiviral",
"responses"
] |
2011
|
NK Cell Receptor/H2-Dk–Dependent
Host Resistance to Viral Infection Is Quantitatively Modulated by
H2q Inhibitory Signals
|
Neurons of the cerebellar nuclei convey the final output of the cerebellum to their targets in various parts of the brain . Within the cerebellum their direct upstream connections originate from inhibitory Purkinje neurons . Purkinje neurons have a complex firing pattern of regular spikes interrupted by intermittent pauses of variable length . How can the cerebellar nucleus process this complex input pattern ? In this modeling study , we investigate different forms of Purkinje neuron simple spike pause synchrony and its influence on candidate coding strategies in the cerebellar nuclei . That is , we investigate how different alignments of synchronous pauses in synthetic Purkinje neuron spike trains affect either time-locking or rate-changes in the downstream nuclei . We find that Purkinje neuron synchrony is mainly represented by changes in the firing rate of cerebellar nuclei neurons . Pause beginning synchronization produced a unique effect on nuclei neuron firing , while the effect of pause ending and pause overlapping synchronization could not be distinguished from each other . Pause beginning synchronization produced better time-locking of nuclear neurons for short length pauses . We also characterize the effect of pause length and spike jitter on the nuclear neuron firing . Additionally , we find that the rate of rebound responses in nuclear neurons after a synchronous pause is controlled by the firing rate of Purkinje neurons preceding it .
Cerebellar nucleus ( CN ) neurons are crucial to the olivo-cerebellar circuit as they provide the sole output of the entire cerebellum [1 , 2] . CN neuron’s firing patterns are of great importance for motor related tasks and representation of movement parameters [3] . Within the cerebellum , their direct upstream connections originate from the Purkinje neurons ( PNs ) . PN firing patterns are vital for CN neuron’s functioning , as CN neurons receive strong inhibition from many PNs [4 , 5] combined with modest depression of the synapse through spillover of GABA from many release sites [6] . PNs exhibit an elaborate firing pattern characterized by simple spikes and complex spikes [7] . Simple spikes are driven by spontaneous intrinsic firing [8 , 9] but are also modulated by excitatory input through the ascending axons and parallel fiber synapses [10] and by feed-forward inhibition through the granule neuron-interneuron-PN pathway [11] . The combination of intrinsic firing and synaptic input results in highly regular spikes with typical short pauses [12 , 13] , where a pause is a short cease in firing . Given this elaborate firing pattern in PNs , how can the downstream CN neuron make sense out of this signal ? Several coding strategies have been proposed for the CN neurons . Broadly , these strategies can be categorized as rate coding or time coding . For a long time it was assumed that CN neurons receive information from PNs by means of a rate code . In cats , CN neurons ( recorded from the anterior interpositus nucleus ) exhibited rate modulations during locomotion [14] . Simultaneous paired recordings from Purkinje and CN neurons show that the modulation responses of the pair is not always reciprocal , which implies that firing modulation characteristics of CN neurons reflect combined activity of many of its presynaptic PNs and mossy fiber input [15] . More recently , time coding was proposed by Person & Raman and entailed time-locking to synchronous input from the PNs [5] . The rationale is that synchrony of a small fraction of PN inputs cause brief periods of relief from inhibition , which makes the CN neuron’s spiking time-locked to the synchronous input . Systematic analysis of simple spikes from neighboring PNs in anaesthetized rats revealed that spikes associated with pauses in firing are far more synchronized ( ±2 ms precision ) than regular firing simple spikes [7] , which indicates a role for pause synchrony in precise timing . The study also reports that , in pairs of PNs , approximately 35% spikes were precisely synchronized and 13% of pauses were synchronized either by their beginning spikes or ending spikes . Moreover , cross-correlograms of neighboring PNs calculated only with pause beginning spikes or ending spikes exhibited sharp peaks and the correlations thus obtained were not distinguishable from each other . So it seems that synchronization of both pause beginning and pause ending spikes are equally probable in neighboring rat PNs . In another study ( also anaesthetized rats ) , timing of pause beginning spikes ( and also ending spikes ) was significantly correlated ( ±5 ms ) in pairs of PNs which exhibited high complex spike synchrony [16] . Coupled pairs of PNs were likely to pause together ( ±5 ms precision ) with an increased probability of 82% and end their pauses together with an increased probability of 39% . In a follow-up review , De Schutter and Steuber proposed that PN simple spike trains combine both rate and temporal codes [17] . The pauses in simple spikes act as a temporal code that may cause a well timed rebound firing of CN neurons when the pauses are synchronized across a population . At the same time , the regular spikes act as a rate code that determines the magnitude of succeeding rebound responses . So it appears that PN synchrony—in either spikes or pauses- may play a role in both time and rate coding in the CN neurons . However , it remains unclear what characteristic of “synchrony” in elaborate PN firing patterns can be represented in the downstream CN . Gauck and Jaeger ( 2000 ) [18] previously investigated the effect of PN synchrony on the rate modulation and spike timing of CN neurons using an experimental dynamic clamp approach . They found that PN synchrony plays an important role in controlling spike timing in CN neurons , but their study assumed that PN spike trains can be described as a Poisson process , which was later shown not to be true [12] . Based on the experimental study of pause synchronization [7] , we defined in this computational work “synchrony” as synchrony of pauses with beginning or ending spikes synchronized , or , as overlapping pauses without spike synchronization . Using experimentally derived parameters for connectivity between the PNs and CN neurons and synthetic PN spike trains , we then investigated how these different types of simple spike synchrony influenced the firing response of the CN neurons . Further , we characterize the effect of PN firing rate preceding the synchronous pause on CN neuron’s firing both during and after the synchronous pause .
The firing of deep CN neurons is highly regular in vitro [18] . In vivo , the firing of these neurons becomes irregular [19] mediated through excitatory and inhibitory synaptic inputs . To match these in vivo data , we tuned both the gain of inhibitory and excitatory synapses impinging on the CN neurons so that the neurons fire spontaneously around 37–38 Hz , which is the observed mean firing rate in vivo of normal rats [19] . The analysis was performed for three model neurons corresponding to three different values of input gain: low , medium , and high , respectively . The gains of excitatory and inhibitory conductance are a specific percentage of their maximal synaptic conductance . The value of inhibitory gain for low , medium and high gain levels was 10% , 70% and 150% , respectively . For each gain level , the excitatory gain value was adjusted so that the model fires at levels of the aforementioned in vivo firing rate [19] . The excitatory gain values corresponding to low , medium and high gain levels were 7% , 12% and 24% respectively . Note that for each value of input gain , the analysis was performed in turn for each of the three models differing in rebound conductance profile ( see Methods ) . Synaptic input had maximum impact on the regularity firing of the neuron at the high gain condition ( Fig 1A–1C ) . A CN neuron’s spike train became more irregular as the gain of the synaptic input increases , as indicated by the increasing coefficient of variation in the inter-spike interval ( Fig 1D ) . As CN neurons exhibit irregular firing pattern in vivo [19] , the high gain condition observed in our study best matches CN firing in awake animals . In order to analyze the influence of PN synchrony on the firing response of the CN neuron , we characterized the neuron’s response to synchronous PN pauses ( see Methods ) . Simulations were run with different synchrony types ( pause beginning , pause ending or pause overlapping ) , with varying amount of input synchronization and different pause length and input gain . This analysis was carried out in all three models differing in rebound conductances ( m1 , m2 , m3 ) . The CN neuron’s response to synchronous pauses of length 20 ms and 40 ms for various values of input gain can be seen in Fig 2 . Fig 3B–3D shows the response of the model neuron to 100 different trials where we quantified the accuracy of time-locking ( Fig 3B ) , extent of increase in rate modulation ( Fig 3C ) , and reliability of increase in firing ( Fig 3D ) for pause beginning synchronization . The figure represents the condition for high gain . Fig 3E–3H illustrates the results of pause overlapping synchronization . Increase in input synchronization resulted in increased rate modulation and an improved reliability of this increase , which will be further analyzed in the subsequent sections . Since most of the results were identical for all the three different rebound models ( Fig 3 ) , the subsequent figures in this study will reflect observations from only one of those models ( m2 ) . We quantified the increase in firing rate of the CN neuron during synchronous pauses with spikes synchronized at the beginning or ending , or overlapping pauses ( where no spikes are synchronized ) , for different amount of synchronization . Synchronous pauses were generated by randomly selecting a pause greater than threshold and temporally aligning it with pauses from other Purkinje neuron spike trains according to the desired amount of synchronization . Pause ending and pause overlapping synchronization consistently mediated greater firing rate increases compared to pause beginning synchronization . This could be seen for both pause values of 20 and 40 ms ( Fig 4 and S1 Fig ) . For all synchronization types , the rate modulation increased with an increase in input gain for all values of synchrony . For example , increase in firing rate produced by 50% and 100% input synchronization increased by approximately 50 Hz and 91Hz respectively , from the low input gain ( Fig 4F ) to the high input gain ( Fig 4H ) condition ( pause value of 40ms , pause beginning synchronization ) . From Fig 4B one can observe that , for both synchrony types a 20 ms pause is too small to mediate any firing rate increase for 25% and 50% of PN synchronization ( low input gain condition ) . Hence for the same conditions , the modulation mediated by these two synchrony types is very small and doesn’t significantly differ from each other ( p>0 . 05 ) . For a pause length of 40 ms , the rate increases exerted in the pause overlapping or pause ending condition is significantly greater than that of pause beginning for all values of synchronization and all values of input gain ( p<0 . 0041 , Fig 4F–4H and S1 Fig ) . The increase in firing rate of the CN neuron exerted by pause ending type synchronization was not significantly different to that of pause overlapping condition for all values of pause length , input gain , and synchronization ( p>0 . 15 ) ( S1 Fig ) . We also simulated the effect of a mixed-type synchronization where PN pauses were synchronized with all three types of synchronization ( beginning , end and overlapping ) in equal proportion . This resulted in increased firing rate modulation of CN neurons , which was greater than that of the pause beginning synchronization , but less than the pause overlapping condition . In order to elucidate the membrane mechanisms underlying increased firing of CN neurons during the synchronous PN pauses , we blocked the rebound conductances ( namely HCN channel , T-type calcium channel and persistent sodium channel ) in the CN neuron model . Fig 5A and 5B shows that when the rebound conductances were blocked , the CN neuron's firing rates within the synchronous PN pause were not altered ( pause beginning synchronization , pause length = 20ms ) . Therefore , the increase in firing rate during synchronous PN pauses is mediated by a passive increase in membrane potential until the CN neuron's firing threshold is reached , at which point fast sodium currents are activated , resulting in spiking activity . The lower increase in firing rates caused by pause beginning spike synchrony is explained by the deep hyperpolarization of the CN neuron’s membrane up to the chloride reversal potential ( at -75 mV ) ( Fig 5C ) , causing the rebound to take longer as compared to the pause ending and pause overlapping conditions . As a result , the spike threshold will be reached later , and the average firing rate measured over the pause duration is therefore lower than that of the other pause synchronization conditions . For all types of synchrony the increase in firing rate was restricted to within the period of synchronous pause and no substantial increase was observed outside the pause period . Next we investigated the effect of the amount of synchrony on the rate modulation of CN neurons . Increase in PN input synchronization resulted in an increased rate modulation for all pause synchrony conditions ( S2 Fig ) . We start by describing the observations related to pause beginning synchronization . Firing rate increase mediated by pause beginning type synchronization increased from 24 . 5 Hz to 109 Hz for 25% to 100% input synchronization respectively ( S2–S2D Fig 40 ms pause ) . For the same synchrony condition and pause length , the amount of synchrony is represented as significant rate increases for medium and high values of input gain ( S2C and S2D Fig p<0 . 0001 ) but not for low input gain ( S2B Fig . p>0 . 1 ) . Input synchrony for the 20 ms pause length is represented as significant rate increases for the high input gain condition ( S2D Fig ) ( p<0 . 0045 ) but not for low and medium gain conditions ( S2B and S2C Fig p>0 . 1 ) . Therefore , the amount of input synchrony is better represented in the firing rate increases of CN neurons by a 40 ms synchronous pause than by a 20 ms synchronous pause . We observed similar results for the pause overlapping synchronization condition: a 40 ms pause represents the amount of presynaptic PN pause synchrony better ( S2E–S2H Fig ) . One notable difference is that input synchrony of the 40 ms pause sequence is significant even for low input gain condition ( S2F Fig p<0 . 005 ) . Thus , we conclude that for both synchrony types the amount of synchrony is better represented in downstream CN neurons when the PN to CN neuron synapses have a high gain because the release from stronger inhibition produces a bigger increase in firing modulation . According to Person and Raman , PN synchronization causes brief periods of relief from inhibition and this time-locks the CN neuron’s response to the input synchronization [5] . This novel form of time coding , which transmits the timing of occurrences of PN synchronization to its downstream targets was suggested to be present both in vivo and in vitro [5] . By varying the amount of pre-synaptic PN synchronization , different pause length ( 20 or 40 ms ) , and synchrony type ( pause beginning or pause overlapping ) , we investigated the effect of pause synchrony on time-locking of the CN neuron spikes . Synchronous pause beginning spikes effectively time-locked the CN neuron’s response to the synchronized event ( Fig 6 ) . For almost all values of input gain and amount of synchronization , the value of vector strength obtained was significant ( p<0 . 01 , Fig 6 ) . Increased input pause synchronization decreased the variability in latency , increased the vector strength and hence the precision of the time-locking phenomenon . For example , for a pause length of 20 ms and pause beginning synchronization , the vector strength for 25% , 50% , 75% and 100% synchronization are 0 . 65 , 0 . 76 , 0 . 85 , 0 . 91 ( p<0 . 01 ) respectively ( Fig 6C1 , high gain condition ) . Brief releases from inhibition produced by the synchronous pause and the deep hyperpolarization of the neuron by the pause beginning spikes caused effective time-locking of CN neuron spikes to the input event . This effect was enhanced at the high input gain condition . Pause overlapping synchronization also time-locked CN neuron’s spiking with better time-locking observed for increased input synchronization . As for pause beginning synchronization , the vector strength for pause overlapping type synchronization was significantly greater than from uniform distribution of spike latencies on a unit circle except for some low gain conditions ( p<0 . 01 Fig 6 ) . Thus brief synchrony of PN pauses time-locks the CN neuron’s spiking . The latency distributions for pause ending synchronization were similar to those for pause overlapping type for all values of gain and pause length ( p>0 . 12 ) ( S3 Fig ) . In the previous section we demonstrated how synchronous PN pauses elicit time-locked CN neuron spikes thereby transmitting the timing information in a reliable manner to downstream targets of CN neuron . We compared the degree of time-locking caused by pause beginning or pause overlapping synchronization by looking at their spike latency distributions in detail . This analysis revealed that time-locking by pause beginning spikes is more accurate than that of pause overlapping condition for a 20 ms ( Fig 7A–7D ) synchronous pause but not for a 40 ms ( Fig 7E and 7F ) pause period . However , this significant increase in the accuracy of time-locking between the two synchronization types for a 20 ms pause sequence is seen only for high gain condition ( Fig 7D , p<0 . 01 , synchronous pause length = 20 ms ) . For low gain the increase in the precision of time-locking between the two types of synchronization is not significant ( Fig 7B , p>0 . 09 ) and for medium gain the increase is significantly seen for 50% , 75% and 100% input synchronization ( Fig 7C , p<0 . 01 ) but not for other values of input synchronization ( p = 0 . 42 ) . With a 40 ms pause the effect of the asynchronous PN inhibition becomes more dominant , resulting in a significant effect for 100% synchronization only . In conclusion , pause beginning type synchronization can cause better time-locking than pause overlapping synchronization depending on the strength and synchrony of inhibition . In order to elucidate membrane mechanisms causing the time-locking of CN neuron spiking to PN pause synchronization , we blocked the rebound conductances in CN neuron model and analyzed its effect on time-locking . Rebound conductances had an insignificant effect on the time-locking of the CN neuron ( Fig 8A and 8B ) . There was no significant change in the variability of the latency when membrane rebound conductances were blocked ( p>0 . 31 ) . Instead , we attribute the time-locking to release from inhibition ( Figs 5C and 5D , 8C and 8E ) , as has been previously demonstrated in other systems [20] . In the presence of strong inhibition , synchronous pause beginning spikes hyperpolarize the neuron's membrane potential and push it to a common state near the chloride reversal potential ( -75 mV ) for every trial ( Fig 8C ) . As a result , the time it takes to ramp up from that potential is always the same and hence precise time-locking is achieved and observed as a narrow distribution of spike latencies , Fig 8D . In pause overlapping and pause ending conditions , the release of inhibition is not orchestrated as such and hence the membrane potential is different across trials , which results in differential times to ramp up to firing threshold . As a consequence , the latency of the first spike is noisier and precise time-locking is reduced ( Fig 8D ) . With low inhibition , even pause beginning synchrony fails to hyperpolarize the membrane potential ( Fig 8E ) to the chloride reversal and spike latencies become noisier , which reduce precise time-locking ( Fig 8F ) . Up to this point , we used precise synchrony of spikes without any jitter . Because this is a rather idealized situation , we also investigated whether jittered pause spikes can still evoke time-locking of CN neuron’s responses . Therefore , we quantified the time-locking behavior of the neuron when presented with synchronous pause differing in the amount of synchronization and pause beginning spikes jittered with various values of jitter . We generated synchronous PN pauses in the same way as described in previous sections but with jittered pause beginning spikes and a pause duration of 40 ms . When the CN neuron's input gain is high , the strong inhibition prevents the neuron to spike during the jitter period . This can be seen in Fig 9A where , for high input gain and 75% PN synchronization , the neuron's first spike post synchronized pause onset is only after the complete jitter period . Also , for the same input condition , CN neuron's first spike latency increased with jitter period , but actually decreased when measured relative to the end of the jittered spikes . For example , the median latency decreased from 13 . 38±1 . 23 ms for 1 ms spike jitter to 10 . 98±1 . 16 ms for 7 ms spike jitter . For the low gain condition , the CN neuron often fired during the jitter period because the inhibition was too weak ( Fig 9B ) . We compared the spike timing accuracy of CN neuron for various values of spike jitter by analyzing the distribution of first spikes within the synchronous pause ( from all trials ) post the jitter period and within the jitter period separately . The variability of spike latency post jitter period was not significantly different from each other when compared between different spike jitter conditions . We observed this for all values of input gain and input synchronization . In contrast , the percentage of trials where the neuron spikes during the jittered period increased as the amount of spike jitter increased for low gain condition . For 75% PN synchronization this value increased from 5% to 15% from a spike jitter value of 3 ms to 7 ms ( Fig 9E ) . Therefore when the strength of PN inhibition is weak , spike-timing precision decreases as the amount of the spike jitter increases . But when PN inhibition is strong , the spike timing precision of CN neuron is the same for all values of spike jitter as the neuron allows its first spike only after the jitter period . Thus , stronger inhibition enables the CN neuron to transmit the timing information accurately even if there is a significant jitter in the incoming spikes . So far our analysis was limited to the effect of synchronous pauses on the firing modulation or time-locking within the pause . Our analysis has revealed that , on average , rebound conductances have no or little effect on this firing modulation or time-locking . CN neurons exhibit rebound firing when the membrane is hyperpolarized for a considerable amount of time and subsequently released from it [21] . De Schutter and Steuber [17] suggested that the amplitude of the CN neuron rebound bursting ( in this study measured as the CN neuron’s firing rate for a duration of 1s from synchronous pause onset ) is directly proportional to the PN firing rates before the synchronous pause . This can be explained by the inactivation/de-inactivation characteristics of rebound conductances ( persistent sodium current , T-type calcium current ) . The rebound conductances that are inactivated during spiking activity of the neuron require substantial membrane hyperpolarization to get de-inactivated [21] . Therefore , high PN firing preceding the synchronous pause leading to more hyperpolarization of the CN neuron's membrane may cause more de-inactivation of rebound conductances and hence higher amplitude of rebound responses of CN neuron . We employed cross-correlation analysis to study the effect on PN firing preceding the synchronous pause on the amplitude of CN neuron's rebound responses . Average PN firing rate ( from all PNs participating in the pause synchrony ) during a period of Δt before the synchronous pause was forced to a particular frequency 'fPN' ( see METHODS ) . We sampled a uniform distribution of fPN from various trials to avoid sampling biases associated with particular PN firing frequencies . CN neuron’s firing rate was quantified for a period of 1s starting from synchronous pause onset . We observed only small increases in the frequency of CN neuron firing in response to increasing average PN firing rate ( Fig 10A–10D ) . Maximum firing rate increase of CN neurons post pause onset increased from 4 Hz to 11 Hz for 25% to 100% input synchronization ( Fig 10A–10D ) . Our results are consistent with that of experiments in mouse cerebellar slices where CN neurons respond to 150 ms , 100 Hz stimulation of presynaptic PN afferents with a moderate post inhibitory firing rate increase of 12 Hz [22] . The absence of larger firing rate increases is due to the limitation imposed by chloride reversal potential of GABAergic currents [22] and the requirement of substantial membrane hyperpolarization to de-inactivate the rebound currents [21 , 22] . The steady state half inactivation value for T-type calcium current and persistent sodium current is around -80 mV [21] . But PN neuron inhibition of CN neurons is limited by vivo chloride reversal potential whose value in CN neuron is around -75 mV [22] . Therefore the rebound conductances exhibit only partial recovery from inactivation during PN inhibition . However , even with this level of recovery from inactivation , the de-inactivation of rebound conductances were directly modulated by the PN firing rate preceding the synchronous pause ( Fig 10E and 10F ) . The greater the firing rate of PNs before the synchronous pause , the greater the recovery from inactivation . Cross-correlation of PN and CN neuron firing rate increase revealed a high significant value for all values of input synchronization . The pearson correlation r increased from r = 0 . 381 in the 25% synchronicity case to r = 0 . 551 in the 100% synchronicity case . Thus the firing rate of PNs before the pause controls the amplitude of succeeding CN rebound firing .
In this study we investigated the effect of PN synchrony in the context of pauses on the spiking response of CN neurons . We stress the importance of pause synchrony in our simulations because the cerebellum is known to control the motor output pattern by transient disinhibition of CN neurons through corresponding pausing of PN firing [23 , 24] . A wide range of motor behavior can be elicited by transient suppression and disinhibition of PN spiking and CN neuron output , respectively [25 , 26] . Further , CN neuron’s spikes are elicited by transient decrease in PN inhibition of minimum 15 ms duration [18] , see also [25] . We find that pause overlapping and pause ending spike synchronies mediate greater firing rate modulation during the synchronous pause than their pause beginning counterparts . It is worth noting that the synchrony types are not uniquely encoded in the firing rate of these neurons as their standard deviations overlap ( Fig 4 ) . Similar to an in vitro study [18] , we also find that the amount of input synchrony modulates the firing of CN neurons with increased input synchrony resulting in an increase in the firing rate of CN neurons . How could the cerebellar system effectively make use of the increase in a CN neuron’s firing rate mediated by the amount and type of PN pause synchronization ? Recently , an in vivo optogenetic study highlighted the importance of input synchrony through transient ‘pausing’ of PNs by photostimulation of molecular layer interneurons expressing channelrhodopsin [25] . They show that various parameters of eye-blink movement kinematics ( amplitude , speed of movement ) could be modulated by the amount of suppression of PN firing ( number of PNs participating in the pause ) and corresponding modulation of firing rates in CN neurons . Therefore , the effect of amount of input synchrony observed in our simulations is directly related to control of movements and behavior in awake , behaving animals . CN neurons exhibit both rate and time coding , but it remains unclear how these neurons can switch between these two different coding schemes . We reasoned that synchronous pause length could play a vital role in controlling the coding strategy of CN neuron . Therefore we compared the effect of pause length , by measuring the effect of synchronous pauses of 20 or 40 ms . For pause beginning synchronization , a 40 ms pause sequence evoked greater firing rate increases than a similar 20 ms pause sequence ( S4F–S4H Fig ) . Conversely , a 20 ms pause sequence was more accurate in transmitting the timing information compared to a 40 ms pause ( S4B–S4D Fig ) . So , pause duration plays a vital role in promoting rate or time coding of the CN neuron with increasing pause length resulting in increased rate modulation and decreased accuracy of time-locking of the CN neuron’s response to a synchronous pause . With respect to time-locking of CN neuron spikes to PN pause synchrony , we found that pause beginning synchronization cause more accurate time-locking than pause overlapping or pause ending type synchronies . A similar phenomenon was reported earlier by Person and Raman [5] where they quantified the extent of time-locking of CN neuron spikes to PN spike synchronization . The time-locking mechanism mentioned in [5] requires synchrony of PN ISIs , the “ISIs” could either be a regular ISI or a pauses whereas the time-locking mechanism quantified in our simulations only deals with synchrony of pauses . Accordingly , the time-locking in our simulations is most likely caused by molecular layer interneurons , which provide powerful inhibition to nearby PNs . Complex spikes can also trigger pauses in simple spike firing of PNs , and complex spike induced pauses can be synchronized across PNs as a single olivary axon targets multiple PNs of the same Aldolase C expression [27] . Nevertheless , PN simple spike pauses outnumber the number of complex spikes by a factor of 25 ( assuming 1 Hz complex spike frequency , 50% PN ISIs contributed by pauses and 50 Hz PN spontaneous firing rate ) [7] . From Fig 3d of [5] one can infer that for any stimulus interval less than 20 ms ( PN firing rates greater than 50 Hz ) , the “brief relief” from inhibition required for time-locking is insufficient so the neuron’s excitability has to be increased ( by applying a DC current ) in order to obtain a response . As such , and in line with our findings , it appears that time-locking is an inherent phenomenon of “longer” length ISIs or so called pauses . Two vital factors affect the precision of the time-locking mechanism: chloride reversal potential and amount of PN pause synchronization . A strongly hyperpolarized chloride reversal potential leads to stronger inhibition and better alignment of the CN neuron repolarization . Similarly , increased PN synchronization also leads to stronger hyperpolarization . The amount of PN synchrony reaching a CN neuron is unknown [1] . This is because PN to CN projection patterns are not completely understood [27 , 28] . Although it is known that CN neurons receive projection from PNs of the same zebrin signature ( PNs in aldolase C positive compartments project to caudoventral CN and those in negative compartments project to rostrodorsal CN ) [28] , little is known about the spatial locations of PNs projecting to specific CN neurons . Anatomical tracing studies indicate that terminal arbors of PNs located in the same aldolase C compartments can be wide , and significantly separated in the downstream CN [28] . Moreover PN axons that are segregated mediolaterally within the same aldolase C compartment may project to non-overlapping areas in the CN [28] . These facts indicate the possibility of CN neurons receiving projections from PNs in multiple longitudinal compartments ( of same aldolase C expression ) . Future research in this area may confirm whether CN neurons receive projections from PNs in single ( same or different lobules ) or multiple rostrocaudally oriented compartments ( of the same aldolase C expression ) . Experiments measuring PN simple spike synchrony within and across various cerebellar regions ( compartments , lobules ) could reveal more information regarding spatial organization of PN synchronization . We conclude that the multiplexed coding scheme [29] of the PN spike train proposed in [17] is preserved in the spiking response of CN neurons to synchronized PN pauses , i . e . , the time locked latency being a timing signal and the CN neuron’s firing rate a rate code .
The CN neuron model used in our study is based on a previously published model [21] . Three different instantiations of the model are used to mimic three different rebound spiking regimes ( m1 , m2 , m3 ) . Rebound responses in [21] were characterized by the presence of fast rebound burst and prolonged rebound spiking activity . m1 is characterized by the presence of fast rebound burst and a pause followed by prolonged rebound period ( Fig 3 , Neuron 1 of [21] ) . m2 is also characterized by the presence of fast rebound burst and a pause but followed by a very strong prolonged rebound period ( Fig 3 , Neuron 2 of [21] ) . Rebound response of m3 consists of fast rebound burst and a transition to prolonged rebound period without a notable pause ( Fig 3 , Neuron 3 of [21] ) . These rebound regimes were implemented in the model through specific combinations of HCN , T-type calcium channel and persistent sodium conductances . Because the results for these three models were nearly identical , we only show results obtained by model m2 . The model rests at a temperature of 32°C . Through out the paper , HCN channel , T-type calcium channel ( CaLVA ) and persistent sodium channel are referred to as rebound conductances . All simulations were performed in NEURON version 7 . 3 [30] . CN neurons receive inhibitory inputs from PNs and excitatory inputs from mossy fibers . Both synaptic pathways are included in our model and described below . Climbing fiber inputs were not simulated . PNs inhibit CN neurons . The number of PN synapses on a single CN neuron is highly debated and values in the literature range from 40 [5] to 860 [4] . We assumed a total of 200 PN connections on a single CN neuron [5 , 6 , 31–33] . These synapses were distributed uniformly on the CN neuron dendritic compartments . Synaptic conductance time course was modeled according to the following equation Gsyn ( t ) = gmax×N×[exp ( −tτdecay ) −exp ( −tτrise ) ] ( 1 ) where τrise and τdecay are rise and decay time constant respectively . gmax is peak synaptic conductance , and N is a normalization factor that makes the maximum of Gsyn ( t ) equal to gmax . PN to CN neuron IPSCs time constants are based on published experimental data [5] . IPSCs reported in this study were recorded at physiological temperatures and exhibit rapid decay with a decay constant of around 2 . 5 ms . We assumed a maximal synaptic conductance gmax = 11 . 7 nS and synaptic reversal potential of -75 mV [34] . The synaptic parameters of PN to CN neuron synapse used in the model are listed in supplementary information S1 Table ) . The rise and decay time constants of PN to CN neuron IPSCs were temperature corrected using experimentally determined Q10 values for synaptic current kinetics ( Q10 = 2 ) [35 , 36] . The PN—CN neuron synapse is characterized by depression of IPSCs [34 , 37] . Synaptic depression reduces the synaptic conductance for subsequent pulses of a high frequency input train . Depression is not instantaneous but follows a time course reaching a steady state depression value . Synaptic depression was modeled by equations described in [12] and steady state depression data taken from multiple pulse depression values from experimental data ( S1 Table ) . The equations for steady state release probability ( Rss ) and time constant of depression ( τ ) was fitted to experimental data [34] . The equations are given below: Rss ( r ) = 0 . 08+0 . 60*exp ( −2 . 84*r ) + 0 . 32*exp ( −0 . 02*r ) ( 2 ) τ ( r ) =2+2500*exp ( −0 . 274*r ) + 100*exp ( −0 . 022*r ) , ( 3 ) where r is the instantaneous frequency . CN neurons are excited by mossy fibers [38] . The mossy fiber to CN neuron synapse is characterized by the presence of both AMPA and NMDA receptors [38] . NMDA receptors of this synapse are characterized by two components: a fast component ( FNMDA ) that shows weak voltage dependence and much slower component ( SNMDA ) that shows strong voltage dependence . NMDA voltage dependence is modeled [21] according to the equation: f ( V ) = 1 ( 1+s1*exp ( −s2*Vm ) ) ( 4 ) AMPA , FNMDA , SNMDA time constants and peak synaptic conductance values , parameters s1 and s2 of voltage dependence were obtained from [38] ( S1 Table ) . We assumed a total of 100 mossy fiber synapses distributed randomly on the dendrites . Mossy fibers in the model were described by Poisson process with a mean firing frequency of 5 Hz , mimicking the background-firing rate of these fibers [39] . AMPA and NMDA synaptic kinetics were temperature corrected using experimentally determined Q10 values for synaptic current kinetics ( Q10 = 2 ) [35 , 36] . PN simple spike trains contain highly regular inter-spike intervals ( ISIs ) and hence cannot be described accurately by a Poisson process [12] . Shin et al [13] showed that regular patterns of PN simple spike ISIs can be described by higher order gamma processes while pause ISIs can be described by lower order gamma processes . Moreover , the estimated orders from regular patterns and pause ISIs followed gamma distributions indicating the presence of multiple processes underlying these two ISI types . Due to the limited availability of experimental spike trains , we amplified the data by generating numerous synthetic PN simple spike trains based on the properties of experimental ones . Synthetic PN simple spikes were generated based on the method described in [12] using spike trains from experimental data [7] . Briefly summarized , synthetic spikes were generated in three steps after deleting the pauses in the simple spikes caused by complex spikes: for each experimental spike train , interspike intervals ( ISIs ) were categorized as regular patterns ( group of regular ISIs ) or pauses based on a measure of short-range variability ( CV2 ) values ( step 1 , S5 Fig ) [11 , 12 , 36] . For each of the regular patterns ( step 2 ) or pauses ( step 3 ) in the experimental spike train , a corresponding regular pattern or pause was generated for the synthetic spike train based on gamma distribution statistics . The similarity between experimental and synthetic ISIs was determined by Kolmogorov-Smirnov test . None of the generated synthetic ISIs was significantly different from experimental ones ( 99% confidence interval , p>0 . 01 ) . S6 Fig compares the distribution of experimental and synthetic ISIs for 9 randomly selected data samples . The distribution of experimental and synthetic ISIs closely matched each other . Pause synchrony is defined by either having the pause beginning or pause ending spike synchronized or no spikes ( overlapping ) synchronized across the population of PN inputs . We introduce synchrony by randomly selecting a pause that was longer than the threshold ( either 20 or 40 ms ) and temporally aligning it with pauses from other spike trains . The pause thresholds were selected so that they are greater than the spontaneous PN simple spike ISIs in vivo [12] . We used two different values of pause threshold so as to compare the effect of pause length on the spiking response of the CN neuron . As the exact amount of PN synchronization reaching a CN neuron is unknown ( see Introduction and Discussion ) , we systematically increased the amount of synchronization in the presynaptic PN pauses and generated sets of spike trains in which 25 , 50 , 75 and 100% of the selected pauses were synchronized . Note that this way the ISIs before and after the pauses are still randomized in accordance to the real spike trains . For simulations where we tested the role of PN firing rate before the pause on CN neuron’s firing rate post pause onset , average PN firing rate during each trial was forced to a particular frequency , ‘fPN' . We used the same number of trials for each fPN . Average PN firing rate for a trial was forced to fPN in the following way: For each PN participating in the synchronous pause , a pause exceeding a threshold of 20 ms and whose preceding spiking for a time period of 100 ms has a firing frequency of fPN , was randomly selected . We then temporally aligned both the pause and preceding spiking with their corresponding counterparts from other PN spike trains . For certain PNs where no desired fPN could be found before a pause , a pause exceeding the threshold was randomly selected . Neuronal spike times were recorded from the simulated CN neurons . The firing rate during synchronous pauses was computed and compared to a control condition ( i . e . , a model run with the same PN and mossy fiber spike sets , but without the synchronous pause ) to investigate changes in the firing rate caused by the release of inhibition . We measured the spike latency with respect to the onset of synchronous pause , the standard deviation/median absolute deviation of which was then used to further assess the degree of time-locking . To avoid the stochastic influences inherent to the synthetic spike trains , we performed 100 or more simulations for each input regime and computed the average change in firing rate , as well as the standard deviation/median absolute deviation of spike latency . We used standard deviation of latency to first spike within the synchronous pause if the distribution of spike latencies of both distributions followed the normal distribution . If either one of the distribution was not normal , the median absolute deviation was used . The normality of the distribution was established by Lilliefors test [40] using 99% confidence interval . A small standard deviation/median absolute deviation means the neuron is firing at almost the same time after a synchronous pause and hence greater accuracy in time-locking or spike timing precision to synchronous pause . We used bootstrapping and one-tailed t-tests to establish the significance and we assumed a confidence interval of 95% . Reliability of increase in firing rate was calculated as percentage of trials showing an increase in firing rate ( Fig 3D ) . Population spike timing histogram ( PSTH ) of all PNs spikes projecting on to CN neuron were constructed by binning the spike times using 1ms bin size , and normalizing it to total number of spikes of the PSTH . The PSTH was constructed using datasets of the 100% PN synchronization ( Figs 3 , 4 and 8 ) condition to indicate the type of synchronization ( pause beginning or ending or overlapping ) . Vector strength [41 , 42] quantifying the degree of time-locking of spikes for various synchronization conditions was quantified in the following way: The spike latency during synchronous pause for each trial was represented as a vector on the unit circle by an angle θ where θ varies from 0 to 2π . Vector strength was computed by performing a vector addition of all vectors from various trials and was normalized to the number of trials . The outcome of this procedure produces a number on [0 , 1] , where 0 means completely random spike times and 1 indicates perfect time-locking of all spikes . Significance of vector strength was established by the Rayleigh statistic [43] . The threshold for vector strength calculated by Rayleigh test is 0 . 2139 . Rayleigh test's critical Z value for 100 trials and 99% confidence interval is 4 . 575 . For simulations where we tested the role of PN firing rate before the synchronous pause on the CN neuron’s firing rate following the immediate pause , the CN neuron’s firing rate was computed for a period of 1s from synchronous pause onset and compared to a control condition ( model run without synchronized spike and pause sequence ) . Linear regression was used to fit a straight line through the data points . Correlation was measured with Pearson’s correlation as r= cov ( X , Y ) σX×σY ( 5 ) where cov ( X , Y ) is the covariance and σ the standard deviation .
|
Neurons can transmit information by two different coding strategies: Rate coding , where the firing rate of the neuron is vital , and time coding where timing of individual spikes carries relevant information . In this study we analyze the importance of brief cessations in firing of the presynaptic neuron ( pauses ) on the spiking of the postsynaptic neuron . We perform this analysis on the inhibitory synaptic connection between Purkinje neurons ( presynaptic ) and nuclear neurons ( postsynaptic ) of the cerebellum . We employ a computational model of nuclear neurons and “synthetic” Purkinje neuron spike trains to study the effect of synchronous pauses on the spiking responses of nuclear neurons . We find that synchronous pauses can cause both well-timed spikes and increased firing rate in the nuclear neuron . In addition , we characterize the effect of pause length , amount and type of pause synchrony , and spike jitter . As such , we conclude that nuclear cells use both rate and time coding to relay upstream spiking information .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Cerebellar Nuclear Neurons Use Time and Rate Coding to Transmit Purkinje Neuron Pauses
|
Synthetic constructs in biotechnology , biocomputing , and modern gene therapy interventions are often based on plasmids or transfected circuits which implement some form of “on-off” switch . For example , the expression of a protein used for therapeutic purposes might be triggered by the recognition of a specific combination of inducers ( e . g . , antigens ) , and memory of this event should be maintained across a cell population until a specific stimulus commands a coordinated shut-off . The robustness of such a design is hampered by molecular ( “intrinsic” ) or environmental ( “extrinsic” ) noise , which may lead to spontaneous changes of state in a subset of the population and is reflected in the bimodality of protein expression , as measured for example using flow cytometry . In this context , a “majority-vote” correction circuit , which brings deviant cells back into the required state , is highly desirable , and quorum-sensing has been suggested as a way for cells to broadcast their states to the population as a whole so as to facilitate consensus . In this paper , we propose what we believe is the first such a design that has mathematically guaranteed properties of stability and auto-correction under certain conditions . Our approach is guided by concepts and theory from the field of “monotone” dynamical systems developed by M . Hirsch , H . Smith , and others . We benchmark our design by comparing it to an existing design which has been the subject of experimental and theoretical studies , illustrating its superiority in stability and self-correction of synchronization errors . Our stability analysis , based on dynamical systems theory , guarantees global convergence to steady states , ruling out unpredictable ( “chaotic” ) behaviors and even sustained oscillations in the limit of convergence . These results are valid no matter what are the values of parameters , and are based only on the wiring diagram . The theory is complemented by extensive computational bifurcation analysis , performed for a biochemically-detailed and biologically-relevant model that we developed . Another novel feature of our approach is that our theorems on exponential stability of steady states for homogeneous or mixed populations are valid independently of the number N of cells in the population , which is usually very large ( N ≫ 1 ) and unknown . We prove that the exponential stability depends on relative proportions of each type of state only . While monotone systems theory has been used previously for systems biology analysis , the current work illustrates its power for synthetic biology design , and thus has wider significance well beyond the application to the important problem of coordination of toggle switches .
We benchmark our design by comparing it to Toggle B2 [4] , see Fig SI-1 . 1 in S1 Text . Despite remarkable properties of design B2 , observed experimentally in controllable experimental settings [4] , and studied theoretically [16 , 17] , the fact that their functional repertoire includes not only a bistable long-term memory but also the generation of stable oscillations suggests that the environment-toggle system must be tightly controlled in order to avoid spontaneous switching , not merely between different expression states but even between different functions . To address this challenge , we propose a novel design , which is endowed with mathematically guaranteed properties of stability and auto-correction . Our approach is closely guided by concepts and theory from the powerful framework of monotone dynamical systems pioneered by M . Hirsch and H . Smith [18–23] . We employ monotone theory to provide guarantees of global convergence to steady states , thus ruling out unpredictable ( “chaotic” ) behaviors and sustained oscillations . These theorems are valid no matter for all values of parameters and are based only on the network structure . We also provide an extensive computational bifurcation analysis of the corresponding biochemically-detailed and biologically-relevant mathematical models . Our results for homogeneous or mixed populations are valid independently of the number of cells in the population ( N ≫ 1 ) , and depend only on the relative proportions of each type of state . As a basic design , we chose a genetic toggle switch consisting of two mutually repressing genes , lacI and tetR [1] . We use two acylated homoserine lactones ( Acyl-HSLs ) , ( i ) N-butanoyl-l-homoserine lactone ( C4-HSL ) secreted by Pseudomonas aeruginosa [24] , and ( ii ) N- ( 3-hydroxy-7-cis-tetradecenoyl ) -L-homoserine lactone ( 3-OH-C14-HSL ) produced by Rhizobium leguminosarum [13] as a means of coordinating toggle-host activity . Our design has two QS arms built-in the toggle in such a way that each promoter-repressor pair is controlled by its own QS signaling pathway symmetrically . Because of this “mirror-like” toggle symmetry , we call our design a symmetric toggle or an “S” design . To benchmark the new S toggle design and the monotone systems approach , we compare the S design to the well-studied asymmetric B2-strain ( Fig SI-1 . 1 in S1 Text ) which has one QS arm only [4 , 16] . In this work , we call the asymmetric B2-strain the “A” design . Our S design cannot be reduced to the A design by removing one QS arm , and , thus , the S design cannot be viewed as a straightforward extension of the A design . From a theoretical standpoint , it is worth remarking that the A design is non-monotone . Under certain experimentally controllable conditions ( e . g , unsaturated levels of AAA+ proteases ClpXP , etc . ) , the S vs . A toggle comparative results obtained in this work can be summarized as follows: In Models , the S toggle and A toggle mathematical models are introduced . The basic formalism and fundamental mathematical results of monotone systems theory , including Strong Monotonicity and Hirsch’s Theorem [18–21 , 23] are as well reviewed . Reference values of dimensionless parameters , scaling , and selection and interpretation of bifurcation parameters are discussed . We also formalize a concept of spontaneous synchronization errors by considering three types of equilibrium populations: One homogeneous population , and two heterogeneous ( mixed ) populations ( bimodal distributions ) with both equally ( 1:1 ) -mixed and not-equally ( N1:N2 ) -mixed transcriptional signatures , N2 ≪ N1 , the latter giving rise to spontaneous synchronization errors , where N = N1 + N2 , and N is the number of cells in the given population . In Results and Discussion , we proceed to a comparative theoretical and computational analysis of the S and A toggle designs . We begin this section with results on the application of monotone systems theory to the S design , as these results constitute the main conceptual and practical subject motivating this work ( Application of Monotone Systems Theory to the S Design ) . We then explain how monotone systems theory allows one to predict , based on qualitative knowledge only , that generically all solutions converge to equilibria , with no possible oscillations [16] nor chaotic behavior [25] , no matter what the kinetic parameters are . This is in contrast to the A design , which may admit oscillations [16] . Next , we analyze single S and A toggles decoupled from the environment ( Bistability in Single S and A Toggles ) , and observe that the S toggle remains bistable even if “redundant” repressor genes are removed from the corresponding plasmids . To show how the S design is more robust than the A design , we carry out a comparative bifurcation analysis of populations consisting of coupled S or A toggles . We select a free ( bifurcation ) dimensionless parameter which can be interpreted in terms of experimental interventions [6] leading to ( a ) changes in the membrane permeability , or ( b ) changes in the half-lives of repressor proteins , or ( c ) changes in the specific growth rate of the host cell . We finally test the toggle design capabilities to self-correct spontaneous synchronization errors by sampling the basin of attractor of the corresponding equilibrium solutions . We find that the S toggle design is able to self-correct synchronization errors far better than the A toggle design . The paper also has Supplemental Information ( SI ) materials S1–S8 which can be found in S1 text ( the file S1-text . pdf ) . In SI-1 Toggle B2 ( see S1 Text ) , we discuss the relationship between A design and its prototype , the E . coli strain “B2” developed by Kobayashi et al [4] who considered a number of genetic toggle switches , interfaced with a QS signaling pathway . In SI-2 Model Derivation ( see S1 Text ) , we derive mathematical models and carry out a nondimensionalization procedure , the conclusions of which are used in the main text ( Scaling ) . In SI-3 Estimation of Parameter Values ( see S1 Text ) , we discuss ranges of biologically meaningful parameter values based on data available in the existing literature . Values of biologically meaningful parameters depend upon experimental conditions and other factors controlled by an experimenter , as reviewed in [6] . Therefore , we provide an example of a concrete estimation of the values of dimensionless parameters , which we interpret in terms of interventions reviewed in [6] . In SI-4 Alternative Definitions of Monotone Systems and Order Preservation ( see S1 Text ) , balanced graphs , relation to graph partitions , and order presentation by flows are explained . In SI-5 Symmetry ( see S1 Text ) , we formalize symmetry of the S design and discuss interpretation of symmetric results with respect to nonsymmetric perturbations typical for experimental systems . In SI-6 Exponential Stability of Cellular Populations ( see S1 Text ) , we prove a number of general theorems to analyze exponential stability [26] of both homogeneous and heterogeneous ( mixed ) population equilibrium states , independently of the number N of cells in the given population , which ( i . e . , the value of N ≥ 2 ) can be a priori unknown . In SI-7 Additional Figures ( see S1 Text ) , we provide additional bifurcation diagrams . In SI-8 Modification of the S and A Models to Describe Sequestration of AAA+ protease ClpXP ( see S1 Text ) , additional ( modified ) mathematical models describing competition of ssrA-tagged protein molecules for AAA+ proteases ClpXP are described .
For the sake of completeness of our description , we begin our discussion of the S toggle and A toggle designs ( Fig 1 ) with two classical orthogonal repressors ( Table 1 ) : Next , the communication network among all toggles ( Fig 1 ) is built upon two quorum-sensing ( QS ) signaling molecules ( Table 1 ) : For the sake of brevity , the QS signaling molecules are called autoinducers G ( C14-HSL ) and R ( C4-HSL ) . Note that the G- and R-signals ( acylated homoserine lactones ) are natural biological signals secreted by Gram-negative bacteria , including E . coli , as a means of coordinating cellular activity [4 , 11] . Finally , to drive the autoinducer concentrations , two synthases are used ( Table 1 ) : Using the above molecular species , we implement and study two different toggle designs called symmetric ( S ) and asymmetric ( A ) designs , respectively , ( Fig 1 ) : The genetic circuit topology used in the A design ( Fig 1 ) is taken from [16] . In order to keep making a fair comparison with the S design , we have replace the luxR-luxI system considered in [16] by the lacI-tetR system suggested in [1] , see SI-1 Toggle B2 in S1 Text . Note that both CinI and RhiI are homologous to LuxI [38] . To host the S and A toggles , we use E . coli , a bacterial cell which has been well-studied in a huge number of relevant experimental and modeling works [32 , 39–50] , and which has been widely used to implement and test various synthetic circuits [1 , 2 , 4 , 7–9 , 15] . A practical modeling reason for this selection is narrowing-down our search for biologically-meaningful parameters to values known from the E . coli studies . However , our conclusions do not depend in any way on biological properties of the host . As a readout of the toggle state in individual cells , we further assume that each E . coli cell contains a gene encoding a spectrally distinct fluorescent reporter , GFP for gene lacI , and RFP for gene tetR , driven by promoters that respond to the autoinducers C14-HSL and C4-HSL , respectively . We do not simulate the processes of bio-synthesis and degradation of the fluorescent proteins explicitly , using appropriate cascade models , for two reasons: ( i ) the “reporter” submodel does not affect the dynamics of the entire model , and ( ii ) the half-lives of the reporter proteins can be made similar to the half-lives of the repressor proteins [2] . Finally , because each toggle can either be in a state where ( a ) LacI protein is abundant , while TetR protein is scarce , or in a state where ( b ) TetR protein is abundant , while LacI protein is scarce , we call state ( a ) a green state or , simply , a G-state and state ( b ) a red state or , simply , an R-state , respectively . Uncertainty about the values of parameters characterizing molecular components of synthetic circuits always presents a significant difficulty in circuit design [2] . Here , we discuss reference values of dimensionless parameters obtained using an appropriate scaling procedure . We also explain how we select and interpret parameters for our bifurcation analysis . A number of powerful concepts and software tools have been developed to efficiently analyze bifurcations of equilibrium solutions in small-scale and medium-sized dynamical models [58–61] . To this end , however , the analysis of bifurcations in the A and S models already becomes a formidable task in terms of CPU loads at N = 10 . For example , the S model describing 10 coupled toggles includes 42 ODEs . Therefore , special conceptual and computational approaches need to be developed to interpret results of modeling with A and S models for cellular populations consisting of thousands or even millions of cells . Fortunately , due to the special structure of the Jacobian matrices for the corresponding linearizations of the A and S models , the computation of the characteristic polynomials , which are used to evaluate stability and bifurcation [26 , 62] , can first be conceptually and , then , numerically simplified , by employing Schur’s formula [63] . As a result , ( i ) the stability and bifurcation analyses of homogeneous populations for any N ≥ 2 can be rigorously reduced to the case of a population consisting of only two toggles , ( ii ) the analysis of a ( 1:1 ) -mixed state for any even N ≥ 4 can be rigorously reduced to the case of only three toggles , and ( iii ) the analysis of a ( N1: N2 ) -mixed state with any N1 ≠ N2 and N1 + N2 = N can be rigorously reduced to the case of only four toggles as described in SI-6 Exponential Stability of Cellular Populations in S1 Text . Schur’s formula [63] also helps to solve another important nontrivial specificity of the A- and S-population models caused by multiplicity of eigenvalues due to the model’s symmetry discussed in SI-6 Exponential Stability of Cellular Populations in S1 Text . Computationally , in the case of multiple eigenvalues caused by symmetry , the standard tools [58–61] cannot be used in a straightforward way , when a special care should be taken . Our theoretic developments can aid in the analysis and interpretation of all such and similar cases arising in modeling of cellular populations , see SI-6 Exponential Stability of Cellular Populations in S1 Text for more rigorous definitions and results . Indeed , the exact ( very large ) number of cells , N , in a cell culture is usually unknown , as cells can die or even be washed out . In such cases , the population density parameter ρ is used , and , therefore , stability of and bifurcation in populations with respect to the variability in their densities is done . The corresponding changes in the integer parameter N that reflect changes in ρ assume a formal study of stability with respect to changes in the number of differential equations in the corresponding models . This is an ill-defined perturbation in the number of equations , and we show how it can be avoided by using the stability approach developed in SI-6 Exponential Stability of Cellular Populations in S1 Text . Capabilities of toggles to fail and recover from spontaneous synchronization errors can be formalized in terms of a multistability concept , that is , as a co-existence of bistable homogeneous populations and various heterogeneous populations ( Fig 2 ) , also called mixed states , under the same conditions . Recall that mixed states are known to lead to bistable distributions [4] . Following [4] , we call a population heterogeneous or , equivalently , mixed if it comprises toggles with different transcription signatures for the same genes: ( i ) the repressor gene lacI is active ( G-state ) , while tetR is repressed , and ( ii ) lacI is repressed , while the repressor gene tetR is active ( R-state ) , see Toggle Designs . In other words , a homogeneous population is fully characterized by either transcription signature ( i ) or ( ii ) , while a heterogeneous population is characterized by mixed signatures ( i ) and ( ii ) simultaneously present in the population ( Fig 2 ) . Different heterogeneous populations can be characterized by transcription signature “mixtures” with ratio ( p:q ) , p + q = 1 , describing the fraction of toggles in the G-state versus the fraction of toggles in the R-state within the same population . For homogeneous populations , we , therefore , have either ( 1:0 ) or ( 0:1 ) transcriptional signature ( Fig 2 ) . With these concepts , we can formulate more precisely our objective: to find conditions under which heterogeneous ( mixed ) population equilibrium solutions can loose their stability or can even be eliminated completely . As a proof of concept , an example of an ( 9:1 ) -heterogeneous population ( Fig 2 ) will be used , where the number of toggles in the first , Green-subpopulation ( G ) ( tetR is suppressed ) is 9 times bigger that the number of toggles in the second , Red-subpopulation ( R ) ( lacI is suppressed ) . In this simplest case , the G-subpopulation comprises 9 cells ( p = 0 . 9 or 90%-fraction of all cells ) , while the R-subpopulation comprises one cell ( q = 0 . 1 or 10%-fraction of all cells ) . Note that our analysis of ( 9:1 ) -mixed states does not depend on the number of cells N in the entire population , which is usually unknown in experiments . In other words , our results hold for any integers N , N1 , and N2 , such that N = N1 + N2 , and N1: N2 = 9: 1 , where the fractions of cells with different transcription signatures are defined by the numbers p = N1/N and q = N2/N , respectively , see SI-6 Exponential Stability of Cellular Populations in S1 Text . The systems considered here are described by the evolution of states , which are time-dependent vectors x ( t ) = ( x1 ( t ) , … , xn ( t ) ) . The components xi represent concentrations of chemical species ( such as proteins , mRNA , metabolites , and so forth ) , the dynamics of which are given by a system of ODE’s: d x 1 d t ( t ) = f 1 ( x 1 ( t ) , x 2 ( t ) , … , x n ( t ) ) , d x 2 d t ( t ) = f 2 ( x 1 ( t ) , x 2 ( t ) , … , x n ( t ) ) , ⋮ d x n d t ( t ) = f n ( x 1 ( t ) , x 2 ( t ) , … , x n ( t ) ) . We also write simply dx/dt = f ( x ) , where f is a differentiable vector function with components fi . The coordinates xi ( t ) are non-negative numbers . We write φ ( t , x0 ) for the solution of the initial value problem x ˙ ( t ) = f ( x ( t ) ) with x ( 0 ) = x0 , or just x ( t ) if x0 is clear from the context , and assume that this solution x ( t ) exists and remains bounded for all t ≥ 0 .
To apply monotone systems theory to the S toggle model Eqs ( 1 ) – ( 6 ) , we first rewrite the model in the following convenient general form with 4N + 2 variables: d x i d t = h x ( x i , y i , g i ) , d y i d t = h y ( x i , y i , r i ) , d g i d t = h g ( y i , g i , g e ) , d r i d t = h r ( x i , r i , r e ) , d g e d t = H g ( g e , g 1 , … , g N ) , d r e d t = H r ( r e , r 1 , … , r N ) . Here , i = 1 , … , N , all the functions in the right-hand side are differentiable , and the following signs hold for partial derivatives , everywhere in the state space: ∂ h x ∂ x i < 0 , ∂ h x ∂ y i < 0 , ∂ h x ∂ g i > 0 , ∂ h x ∂ a 1 > 0 , ∂ h x ∂ a 3 > 0 , ( 22 ) ∂ h y ∂ x i < 0 , ∂ h y ∂ y i < 0 , ∂ h y ∂ r i > 0 , ∂ h y ∂ a 2 > 0 , ∂ h y ∂ a 6 > 0 , ( 23 ) ∂ h g ∂ y i < 0 , ∂ h g ∂ g i < 0 , ∂ h g ∂ g e > 0 , ∂ h g ∂ a 5 > 0 , ∂ h g ∂ δ < 0 , ( 24 ) ∂ h r ∂ x i < 0 , ∂ h r ∂ r i < 0 , ∂ h r ∂ r e > 0 , ∂ h g ∂ a 6 > 0 , ∂ h r ∂ δ < 0 , ( 25 ) ∂ H g ∂ g i > 0 , ∂ H g ∂ g e < 0 , ∂ H g ∂ δ e < 0 , ( 26 ) ∂ H r ∂ r i > 0 , ∂ H r ∂ r e < 0 , ∂ H r ∂ δ e < 0 , i = 1 … , N . ( 27 ) Next we observe that the S system is monotone , because we may partition its state variables as follows . One set consists of x i , g i , g e , i = 1 , . . . , N , ( 28 ) and another set consists of y i , r i , r e , i = 1 , . . . , N . ( 29 ) Moreover , the corresponding graph is strongly connected , as we have the following paths , for each two indices i , j: x j ⊣ r j → r e → r i → y i ⊣ g i → g e → g i → x i ( 30 ) which shows that one can reach any node from any other node by means of a directed path . Thus , the S model Eqs ( 1 ) – ( 6 ) is strongly monotone . We conclude as follows . Theorem 1 Typical solutions of the S model Eqs ( 1 ) – ( 6 ) converge to steady states . This fundamental result is robust to parameters as well as to the functional form of the equations . It ensures that our proposed design has theoretically guaranteed global stability properties . No stable oscillations [16] can exist , nor can other ( for , example , “chaotic” [25] ) solution regimes arise . In addition to these global properties , it is also possible to use the theory of monotone systems in order to make qualitative predictions about bifurcation diagrams as discussed in the next section . The monotonicity property of the S system has important consequences regarding its transient as well as asymptotic behavior . We discuss in an appendix how Kamke’s Theorem characterizes order relations for monotone systems . We explain now what these mean , explicitly , for the S system . Let zi ( t ) characterize the state of the i-th S toggle at time t ≥ 0 , that is , zi ( t ) = ( xi ( t ) , yi ( t ) , gi ( t ) , ri ( t ) ) , i = 1 , … , N . Let Z ( t ) characterize the state of the population of cells , Z ( t ) = ( z1 ( t ) , … , zN ( t ) , ge ( t ) , re ( t ) ) . Suppose that we have two initial sets , Z ( 0 ) and Z ˜ ( 0 ) , of values for the various expression levels of the repressor proteins , LacI and TetR , and we consider the behavior of Z ( t ) and Z ˜ ( t ) for t > 0 . Now suppose that we wish to understand what is the effect of a perturbation in one of the components of the initial state zi ( 0 ) for S toggle i with some fixed i , 1 ≤ i ≤ N . ( A similar argument can be applied to perturbations in other components of the initial state , or even simultaneous perturbations in all the components . ) Suppose , for example , that we are interested in understanding the behavior starting from a state in which x ˜ 3 ( 0 ) ≥ x 3 ( 0 ) in the 3rd toggle z3 . This gives rise to a new population-wide solution Z ˜ ( t ) , and we use a tilde to denote its coordinates , that is , Z ˜ ( t ) = ( z ˜ 1 ( t ) , … , z ˜ N ( t ) , g ˜ e ( t ) , r ˜ e ( t ) ) , where z ˜ i ( t ) = ( x i ( t ) , y i ( t ) , g i ( t ) , r i ( t ) ) , i = 1 , … , N . Then , using the information provided by the partition shown in Eqs ( 28 ) and ( 29 ) , we can predict that , for all t > 0: x ˜ i ( t ) ≥ x i ( t ) , y ˜ i ( t ) ≤ y i ( t ) , g ˜ i ( t ) ≥ g i ( t ) , r ˜ i ( t ) ≤ r i ( t ) , g ˜ e ( t ) ≥ g e ( t ) , and r ˜ e ( t ) ≤ r e ( t ) for all i = 1 , … , N . As we will see shortly below , a similar conclusion can also be made with respect to perturbations in parameters , not merely initial states . As a first step , we can include the eight parameters , ai ( i = 1 , … , 6 ) , δg , and δr , as constant state variables by formally adding the corresponding equations dai/dt = 0 ( i = 1 , … , 6 ) , and dδg/dt = dδr/dt = 0 to the S-model Eqs ( 1 ) – ( 6 ) . The extended S-model is a monotone system . However , this extended model has no strong monotonicity property , because the nodes corresponding to the parameters cannot be reached from other nodes , as the parametric extension violates the strong connectivity relationships Eq ( 30 ) . However , this is not of any consequence , as the global stability properties of the S system are determined by constant values of the parameters . We only introduced the extended model in the context of bifurcation analysis . One might add additional constant variables to represent other parameters , such as the d’s . These other parameters do not lead to monotonicity , and this lack of monotonicity will have important consequences in bifurcation analysis , as we discuss later . Dependencies between the S-model state variables and parameters Eqs ( 22 ) – ( 27 ) are shown in Fig 3 ( Top Panel ) . Here , the set of all molecular entities in the S design is partitioned into two “orthogonal” subsets , S− and S+ ( Definition of monotone systems ) . Solid arrows and lines highlighted in light brown color correspond to S− , while solid arrows and lines highlighted in cyan color correspond to S+ . Although interactions within each subset contribute to its activate state , the orthogonal subsets repress one another . Here , ClpXP is a pool of AAA+ proteases ClpXP that use the energy of ATP binding and hydrolysis to perform mechanical work during targeted protein degradation within the cell . The corresponding inhibitory ( degradation ) interactions are shown , using dashed gray lines . If the circuit operates near the saturation condition for the pool of AAA+ proteases ClpXP , the S design may loss its monotone properties . However , there is a substantial body of literature that gives theorems to the effect that “small” perturbations of monotone systems retain the properties of monotone systems , for example , a smooth regular perturbation of a strongly monotone system also has generic convergence properties [65] . A similar result as well holds for singular perturbations [69] . An example of three identical S toggles interacting via common autoinducers and operating far from the saturation of AAA+ proteases ClpXP is shown in Fig 3 ( Bottom Panel ) . The monotonicity of the extended model implies that stable loci in bifurcation diagrams depend monotonically on parameter variations . They will increase when the parameter being varied belongs to the component as the variable being analyzed , and will decrease if they are in different components . This property is a consequence of the general order preserving properties of monotone systems , as we explain now . Suppose that x ¯ 0 is a steady state corresponding to a parameter value p0 , that is to say , f ( x ¯ 0 , p 0 ) = 0 . Suppose that we now consider p1 that is very close to p0 and larger than p0 , p1 > p0 . Suppose in addition that x ¯ 1 is a steady state for the parameter value p1 , f ( x ¯ 1 , p 1 ) = 0 , and that x ¯ 1 is stable . Now pick the solution x1 ( t ) of x ˙ = f ( x , p 1 ) that has initial condition x 1 ( 0 ) = x ¯ 0 . Suppose that the extended system x ˙ = f ( x , p ) and p ˙ = 0 is monotone . Now , we may consider the following two initial states for the extended system: ( x ¯ 0 , p 0 ) and ( x ¯ 0 , p 1 ) . Since the second state is larger ( in the sense of Kamke’s Theorem as earlier explained ) in the monotone order , it follows that the solutions satisfy x 1 ( t ) ≥ x ¯ 0 for all t > 0 , and therefore , taking limits , we conclude that x ¯ 1 > x ¯ 0 , as desired . Using Fig 3 in conjunction with the dimension analysis in terms of the relationships Eqs ( 18 ) , ( 19 ) , ( 20 ) and ( 21 ) , certain qualitative predictions can be made about the parametric dependencies based on monotone systems theory . To benchmark the approach , we have selected , as an example , a subset of dependencies shown in Fig 3 , presented in Fig 4A and 4B . Fig 4C and 4D correspond to the case when the S toggle operates under the saturation condition for the pool of AAA+ proteases ClpXP ( SI-8 Modification of the S and A Models to Describe Sequestration 617 of AAA+ protease ClpXP in S1 Text . ) In Fig 4 , three different stable populations are chosen: ( 1 ) an G-homogeneous population; ( 2 ) an ( 1:1 ) -mixed population; here , the levels of LacI and C14-HSL from one subpopulation ( within which LacI is over-expressed ) are shown; and ( 3 ) a ( 9:1 ) -mixed population ( a spontaneous synchronization error ) ; here , again , the levels of LacI and C14-HSL from the largest subpopulation ( within which LacI is over-expressed ) are shown . Because the stable mixed populations do not exist for large values of the parameter d in the cases shown in Fig 4A and 4B , we use both d = 0 . 1 ( weak coupling ) for all populations and , additionally , we use d = 10 ( strong coupling ) for the G-homogeneous population only . In the cases shown in Fig 4C and 4D , the mixed populations turn out to be more robust and exist at d = 10 . Using the S-model Eqs ( 1 ) – ( 6 ) and its sequestration version ( SI-8 . 1 ) ( see SI-8 . 1 Modification of the S Model in S1 Text ) with the values of fixed parameters given in Eqs ( 11 ) – ( 14 ) and ( 15 ) – ( 17 ) , respectively , we find that Fig 3 predicts monotonically increasing dependencies . The loss of stability and disappearance of the mixed states shown in Fig 4B and 4D as a5 increases can be interpreted intuitively by the fact that an increase in a5 leads to an increase in the intracellular levels of the corresponding QS signaling molecules , which , in turn , lead to an increase of extracellular levels of the QS molecules via diffusion , thereby facilitating self-synchronization of the given population of all toggles under conditions corresponding to a stronger interaction among all toggles . In particular , the strong interaction and coupling condition eliminates spontaneous synchronization errors in terms of suppressing the emergence of undesired ( 9:1 ) -mixed states . This result is similar to a well-known fact for oscillators coupled via a common medium that a transition from an unsynchronized to a synchronized regime emerges as the strength of coupling increases [15–17 , 25 , 56] . Indeed , many microbial species accomplish this via quorum sensing , which entails the secretion and detection of diffusible molecules ( autoinducers ) , whose concentration serves as a proxy for population density [10] . Using the expression for the dimensionless parameter a5 given in Eqs ( 18 ) and ( 19 ) , see Scaling , we can conclude that the increase in the values of the parameter a5 leading to the bifurcation point LP2 ( Fig 4 ) can be achieved by the following experimental interventions: We have used bifurcation analysis with respect to changes in the values of the parameter a5 as a way to illustrate predictions from monotone systems theory , and in the process we obtained conclusions regarding improvements of S toggle self-synchronization properties by eliminating the ( 9:1 ) -mixed state . To this end , we note that there is no need to further increase values of a5 to move the system to the bifurcation point LP1 at which the ( 1:1 ) -mixed state loses it stability and disappears , because we do not interpret the ( 1:1 ) -mixed state as a spontaneous synchronization error , see Spontaneous Synchronization Errors . Additional parametric dependencies with respect to changes in other parameters are shown in Figs SI-7 . 1 and SI-7 . 1 in S1 Text . We then repeat the analysis of the same parametric dependencies for a ( 1:1 ) -mixed state , illustrated in Fig 5 and Fig SI-7 . 3 in S1 Text . Like in the previous case , we observe that all dependencies are in line with the predictions suggested by Fig 3 . To this end , we will not provide here reproduced similar results for the saturated S design ( SI-8 . 1 Modification of the S Model in S1 Text ) , because in all computationally investigated cases , the parameter monotone dependencies are predicted by the theory and Fig 3 . The LP-bifurcation point ( Fig 5 ) can be interpreted as follows . Decreasing values of both parameters δg and δr leads to an increase in the intracellular and extracellular levels of the corresponding QS signaling molecules , which , in turn , leads to stronger interactions among all toggles . Indeed , it follows from Eqs ( 20 ) and ( 21 ) ( see Scaling ) that the described changes in the values of dimensionless parameters δg and δr can be achieved by increasing half-lives of the corresponding QS signaling molecules . To this end and similarly to the interpretation provided earlier , as the values of the parameters δg and δr decrease , the ( 1:1 ) -mixed state loses its stability and disappear via an LP-bifurcation ( Fig 5 ) , the effect which is similar to the well-known fact that oscillators coupled via common medium synchronize as the strength of coupling increases [15 , 25 , 56] . We note that the parametric dependencies for unstable solutions are not described by Fig 3 . To explain this observation , we recall that our proof of monotone dependence on parameters applies to stable solutions only , see above . Finally , the monotone parametric dependencies for ( 9:1 ) -mixed states corresponding to spontaneous synchronization errors are illustrated in Figs SI-7 . 4 and SI-7 . 5 in S1 Text . In this case , by increasing the strength of interactions between the toggles from the large subpopulation , the spontaneous error can also be eliminated , corresponding to the existence of the LP-points in panels ( A ) and ( B ) of Figs SI-7 . 4 and SI-7 . 5 in S1 Text . At the same time , increasing the strength of interactions between the toggles from the small population , the corresponding spontaneous error cannot be eliminated . Before comparing population properties of our S design to those of the A design , we remark that , even for isolated cells ( when the diffusion constant d is zero ) , there is a larger range of bistability for the S design compared to the A design . Specifically , a bistability region for a single A toggle in the plane ( a1 , a2 ) at d = 0 is shown in Fig 6 ( top panel ) . Similar regions were found in [1 , 16] . We also observe that the entire quadrant , a1 ≥ 0 and a2 ≥ 0 , spans a bistability region for the S-model at the fixed parameter values given in Eqs ( 11 ) – ( 14 ) . We have computed the bistablility regions for the S design for three different nonzero values of the promoter leakiness parameter γ = 0 . 01 , 0 . 1 , 1 . 0 , respectively , while all other parameter values were kept fixed as in the reference set Eqs ( 11 ) – ( 14 ) , and found that in all the three cases , the entire quadrant , a1 ≥ 0 and a2 ≥ 0 , belongs to the computed regions . In contrast , the bistability region for the A design depends on the promoter leakiness parameter significantly [16] , and we also observed computationally that the bistability region was leaving the domain shown in Fig 6 ( top panel ) as soon as γ was allowed to take on values larger than 0 . 5 , that is , when γ > 0 . 5 . Another important observation that follows immediately from Fig 6 ( top panel ) is that in the case of the S toggle , bistability persists at the origin of the non-negative quadrant in the plane ( a1 , a2 ) , that is , at a1 = a2 = 0 . The observation remains true even for the nonzero values of the leakiness parameters as discussed earlier . Additionally , the property persists for the saturated S design ( SI-8 . 1 Modification of the S Model in S1 Text ) with the updated parameter set Eqs ( 15 ) – ( 17 ) . This simply means that the genes lacI and tetR can be removed from the corresponding plasmids bearing promoters PY and PX , respectively ( Fig 1 ) . In this case ( Fig 6 ) ( bottom panel ) , it is enough to keep the genes on the plasmids bearing the corresponding promoters PG and PR ( Fig 1 ) . We view the reduced S toggle as a minimal design that could be implemented experimentally . The fuller construct S is interesting too , in so far as it is based on the well-characterized and studied Cantor-Collins switch , coupled to quorum-sensing components [4] . We find that the full and reduced designs do not differ much in performance , and , so , we do not consider the minimal design in the rest of the paper . Bistable homogeneous populations of S toggles persist within large ranges of the model parameters . For example , panels ( A ) and ( B ) in Fig SI-7 . 6 in S1 Text show scaled levels of LacI and C14-HSL , respectively , for a homogeneous population of S toggle in the G-state , depending on the values of the diffusion ( membrane permeability ) parameter d . Panels ( C ) — ( F ) in Fig SI-7 . 6 in S1 Text show two stable homogeneous populations of A toggle which coexist while the parameter d is allowed to vary . Because the A toggle design does not have any intrinsic symmetry , the levels of the activated repressor proteins , LacI for the G-homogeneous population shown in panels ( C ) and ( E ) , and TetR for the R-homogeneous population shown panels ( D ) and ( F ) , differ significantly from one another . Recall that the levels of LacI and TetR in the corresponding G- and R-homogeneous populations consisting of S toggles are identically the same due to mirror symmetry . Our intensive computational studies confirm that the discussed results on the stable homogeneous populations of S and A toggles , as well as their dependencies on the diffusion parameter d , are robust with respect to perturbations in the model parameters , including various combinations in the values of the Hill coefficients , promoter leakiness , and the saturation conditions ( SI-8 Modification of the S and A Models to Describe Sequestration of AAA+ protease ClpXP in S1 Text ) . The combination of the analyses discussed here can be summarized by saying that under each one of the two designs , S and A , including biological variability in the Hill coefficients , promoter leakiness , and the degradation sequestration conditions , bistable homogeneous stable populations are possible , in either “Red” or “Green” consensus states , and with the same order of magnitude of expression . The difference between these designs , including the sequestration effect for AAA+ proteases ClpXP , become evident , when we study heterogeneous ( mixed ) populations , as discussed next . Fig 7 shows richness of dynamic effects ( bifurcations ) for a ( 1:1 ) -mixed population of S toggles . We see that as soon as the parameter d takes on larger values , the ( 1:1 ) -mixed state loses its stability via a Branch Point ( BP ) bifurcation [62] ( alternatively called “pitchfork” or “symmetry-breaking” bifurcation [70 , 71] ) , giving rise to two stable ( 1:1 ) -mixed non-symmetric states at d ≈ 1 . 43 . The general symmetry-breaking phenomenon is rigorously studied in [72 , 73] . The symmetry-breaking scenario can be described intuitively as follows . Suppose that we start with a mixed population in which 50% of the cells are in “green” state and 50% of the cells are in “red” state , and the nondimensional diffusion coefficient d ( which , as we saw , in fact incorporates many of the kinetic parameters in the original system ) has a low value . Suppose that we now slowly increase the value of d , and ask what happens to the ( 1:1 ) -mixed state . The first event that is observed , at d ≈ 1 . 43 corresponding to the BP points in all panels of Fig 7 , is that this “pure 50–50 mixed state” loses its stability . A new mixed state arises ( Fig 8 ) , in which there are two subpopulations , one in which green gene-expression dominates ( but with different expression levels of LacI in each of them ) , and another one which red gene-expression dominates ( also with different TetR levels ) . These two mixed states correspond to the solution branches connecting points marked with labels BP and upper LP , and BP and lower LP , respectively , shown in all panels of Fig 7 . Furthermore , as d is increased a bit more ( past d ≈ 2 . 07 corresponding to the two points labeled with LP in all panels of Fig 7 , respectively ) , even these mixed states disappear ( Fig 7 ) . Thus , even with moderate diffusion , heterogeneous populations cannot be sustained , emphasizing the consensus-forming character of the S design . This is in marked contrast to the A design , as shown next . The loss of stability by the ( 1:1 ) -mixed state increases the robustness of the S toggle design towards its self-synchronization by reducing the number of alternative stable states to which the toggle state can settle . In contrast to ( 1:1 ) -mixed populations of ( unsaturated ) S toggles described by the S model Eqs ( 1 ) – ( 6 ) , we observe from Fig SI-7 . 7 in S1 Text that the original ( 1:1 ) -mixed A-population cannot be eliminated ( made unstable ) by increasing the values of the parameter d within a very large parameter interval . In other words , increasing the strength of interactions between the cells does not help to establish synchronization across the given population of identical A toggles . This is in a total agreement with a similar observation reported in [16] , where the A model is studied in great detail . Specifically , it is found that a strong interaction between A toggles ( e . g . , high permeability of the membrane to the autoinducer similar to higher values of d ) results in the suppression of synchronous oscillations , leading to a transition of the population to a stable heterogeneous state , where individual A toggles are locked in different equilibrium states . Our computational experiments with ( 1:1 ) -mixed populations of ( saturated ) Sm toggles described by the Sm model ( SI-8 . 1 ) ( see SI-8 . 1 Modification of the S Model in S1 Text ) led to dependencies qualitatively indistinguishable from those shown in Fig SI-7 . 7 in S1 Text . Therefore , we can conclude that the degradation saturation ( sequestration ) effect may prevent the elimination of the undesired mixed states and synchronization . Next , we consider bistable ( 9:1 ) -mixed populations of S Toggles , which as discussed in the introduction , we think of as arising from random synchronization errors . We observe that ( 9:1 ) -populations of S toggles become quickly extinct as soon as the values of the nondimensional diffusion parameter d are slightly increased ( Fig 9 ) . In contrast to the S design , in the A design , the mixed ( 9:1 ) - and ( 1:9 ) -heterogeneous populations that might arise from random state switching cannot be eliminated by changes in the values of the parameter d ( Fig SI-7 . 8 in S1 Text ) . This is again in a total agreement with a similar observation reported in [16] . Using simulations carried out with the Sm model ( SI-8 . 1 Modification of the S Model in S1 Text ) , we also observed that the sequestration effect results in stable ( 9:1 ) -mixed states for the S design existing for large ranges of the diffusion parameter d . To probe and compare capabilities of the S toggle and A toggle designs to correct “spontaneous synchronization errors” caused by a random flip of one toggle ( or a small fraction of toggles ) from a homogeneous population to the state opposite to the transcription signature adopted by the majority of the cells , we have performed simple random tests . In mathematical and computational terms , these random tests can be interpreted as an elementary numerical procedure to evaluate the size of the basin of attraction for the corresponding equilibrium solutions by sampling the corresponding small neighborhoods of the solutions , using random initial conditions , for each parameter value d ∈ {0 . 01 , 10 , 100} as follows , ( 1 ) find stable G- and R-homogeneous states ( for any population size ! ) , ( 2 ) flip 10% of population , and ( 3 ) explore initial conditions in neighborhood of this state value for the corresponding state variable ( for the S design , since symmetric , only the G-homogeneous state needs to be explored ) . We can conclude from Fig SI-7 . 9 in S1 Text that the A toggle does not have any capability for self-correction of spontaneous errors for all tested values of the parameter d ( Fig SI-7 . 9 in S1 Text ) . The S toggle can self-correct spontaneous synchronization errors for the medium and large values of the parameter d ( Fig 10 ) for all parameters values for which the mixed state becomes unstable , see Fig 9 ( ( 9:1 ) -Mixed Population of S Toggles . ) The rate of the error correction can be to some extend characterized by the observation that the error is corrected within the first 10 minutes counted from its onset ( Fig 10 ) . Unfortunately , theory does not preclude damped oscillations . Thus , all we can do is to make computational estimates in realistic parameter ranges . To check how the limited availability of AAA+ proteases ClpXP may negatively impact the self-correctness property by the S design , we have developed an additional Sm model describing saturation ( sequestration ) of the AAA+ proteases ClpXP ( SI-8 . 1 Modification of the S Model in S1 Text . ) We then conducted additional computations to show that the sequestration , while preserving monotonicity and bistability properties , can lead to the loss of the self-correction of spontaneous errors by the S design . Thus , high levels of these proteases are required to implement successfully the S design . Finally , we note that the reported results on the ( 9:1 ) -mixed states for both S and A designs are independent of the number N of cells in the given population with density ρ and can be applied to any population consisting of thousands or even millions of cells , split into two subpopulations comprising 90% and 10% fractions of all cells with different transcription signatures , respectively ( Stability and Bifurcation in Cellular Populations . ) Specifically , if the given ( 9:1 ) -mixed state is unstable for the S design in the model of 10 ( identical ) cells , it will be unstable in the model describing a larger population of ( identical ) cells because its stability is determined form an auxiliary system of four cells only ( SI-6 Exponential Stability of Cellular Populations in S1 Text . ) The same is true for the stable ( 9:1 ) -mixed state for the A design . In this study , we have shown how synthetic bistable circuits ( toggles ) , and hosting them , programmable cellular populations , can be designed so as to solve a robust molecular task , the maintenance of a coordinated state , and a “majority-vote” auto-correction of deviations , of a binary switch . Our design was guided by concepts from monotone systems theory [18–23] . Specifically , we have shown how this concept can be used for the design of a new class of monotone synthetic biological toggles , including predictive capabilities describing both dynamic state variables and monotone parametric tendencies caused by parameter perturbations . To benchmark the new toggle design , termed the S design , and the monotone systems approach , we have compared the S design with the known ( and non-monotone ) B2-strain from [4] , termed the asymmetric or A design in this work . The B2-strain has been previously studied both experimentally [4] and theoretically [16 , 17] . Despite a number of remarkable properties of the B2-strain ( A design ) , the A toggle multifunctionality suggests that the design must be tightly controlled to avoid spontaneous switching not only between different expression states , but , as well , between different functions such as a bistable memory and an oscillatory phenotype . In this respect , modern gene therapy interventions are currently limited to transfected genes to be either in an “on” or “off” state , when the expression of the transfected gene needs to be regulated tightly for the effective treatment of many diseases . To address this challenge , the monotone S toggle design completely excludes any unpredictable chaotic behaviors , as well as undesired stable oscillations . This conclusion is valid ( of course , under certain experimentally controllable conditions pointed out in this work ) for all parameter values , and provides a strong theoretical guarantee missing from other synthetic biology designs . Some of conditions include: ( i ) a reduced promoter leakiness [51] , and ( ii ) unsaturated levels of AAA+ proteases ClpXP . To achieve an in-depth understanding of dynamic properties of the S toggle design , we have developed biochemically-detailed and biologically-relevant mathematical models to test predictions of monotone systems theory by employing computational bifurcation analysis . To have all results biologically grounded , concrete molecular entities have been used , though the results are general and independent of any specific details . To investigate the effect of a spontaneous toggle switching within cellular populations , leading to bimodal distributions , we have formalized a concept of spontaneous synchronization errors and tested the toggle design capabilities to self-correct spontaneous synchronization errors by sampling the basin of attraction of the corresponding equilibrium solutions . We found that the S toggle design was able to self-correct ( or , auto-correct ) synchronization errors , while the non-monotone A toggle design was not . Because the number of cells in populations is a priori unknown , all the above results and conclusions can make sense only if they are made independently of the population size . To justify the above assertion , we have proved a few general theorems on the exponential stability of the equilibrium solutions corresponding to both homogeneous and mixed populations . The simple exponential stability results are independent of the number of cells in the populations and are based on basic first principles of stability analysis resulting from the Schur’s formula [63] , allowing the characteristic polynomials for the corresponding model linearizations to be computed explicitly . Using an additional model describing saturation ( and sequestration ) of AAA+ proteases ClpXP ( SI-8 Modification of the S and A Models to Describe Sequestration of AAA+ protease ClpXP in S1 Text ) , we have observed computationally that even when the above-mentioned conditions ( i ) and ( ii ) are violated , the S toggle still demonstrates the monotonicity properties . If proteases are in limited supply , however , the conclusions break down , because of the non-monotonicity arising from resource competition . Thus , an important consideration when practically implementing our design is to express these proteases at a high enough level . We remark that our design is based on a bistable design based on deterministic models . This approach is normally used in synthetic biology design of toggle switches , and our goal was to employ ready technology . On the other hand , in gene regulatory networks bistability , or , to be more precise , multi-modality of steady state distributions , may arise in deterministically monostable systems due to low molecule number effects . Intuitively , a slow switch between two promoter states ( modeled in simplest terms by a two-state Markov chain ) gives rise to a “bimodal” distribution of gene activation ( gene is “on” or “off” ) ; this process then may drive a large-molecule number mRNA and protein process , in effect creating a bimodal protein distribution , though this bimodality would be “averaged out” in a deterministic model that considers a large population . In this context , one may mention the work by Thomas et al . [74] which provides a system with two mutually repressing promoters using noncooperative transcriptional regulation but supplemented by a translational control component in which the protein product of one gene binds and degrades the mRNA of the other gene . Because we use cooperative binding ( Hill coefficients 2 and larger ) , our design is specifically geared to bistability even in low noise situations , and the engineered consensus mechanism is designed to correct for noise-induced transitions . It would be interesting in further work to study consensus designs for toggles based on stochastic bimodality .
|
For the last decade , outstanding progress has been made , and considerable practical experience has accumulated , in the construction of elementary genetic circuits that perform various tasks , such as memory storage and logical operations , in response to both exogenous and endogenous stimuli . Using modern molecular “plug-and-play” technologies , various ( re- ) programmable cellular populations can be engineered , and they can be combined into more complex cellular systems . Among all engineered synthetic circuits , a toggle , a robust bistable switch leading to a binary response dynamics , is the simplest basic synthetic biology device , analogous to the “flip-flop” or latch in electronic design , and it plays a key role in biotechnology , biocomputing , and proposed gene therapies . However , despite many remarkable properties of the existing toggle designs , they must be tightly controlled in order to avoid spontaneous switching between different expression states ( loss of long-term memory ) or even the breakdown of stability through the generation of stable oscillations . To address this concrete challenge , we have developed a new design for quorum-sensing synthetic toggles , based on monotone dynamical systems theory . Our design is endowed with strong theoretical guarantees that completely exclude unpredictable chaotic behaviors in the limit of convergence , as well as undesired stable oscillations , and leads to robust consensus states .
|
[
"Abstract",
"Introduction",
"Models",
"Results",
"and",
"Discussion"
] |
[
"infographics",
"chemical",
"compounds",
"engineering",
"and",
"technology",
"enzymes",
"synthetic",
"biology",
"enzymology",
"lactones",
"mathematical",
"models",
"esters",
"dna",
"transcription",
"systems",
"science",
"mathematics",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"gene",
"expression",
"chemistry",
"dynamical",
"systems",
"biochemistry",
"data",
"visualization",
"graphs",
"genetics",
"biology",
"and",
"life",
"sciences",
"proteases",
"physical",
"sciences"
] |
2016
|
Quorum-Sensing Synchronization of Synthetic Toggle Switches: A Design Based on Monotone Dynamical Systems Theory
|
The capacity to respond to day length , photoperiodism , is crucial for flowering plants to adapt to seasonal change . The photoperiodic control of flowering in plants is mediated by a long-distance mobile floral stimulus called florigen that moves from leaves to the shoot apex . Although the proteins encoded by FLOWERING LOCUS T ( FT ) in Arabidopsis and its orthologs in other plants are identified as the long-sought florigen , whether their transport is a simple diffusion process or under regulation remains elusive . Here we show that an endoplasmic reticulum ( ER ) membrane protein , FT-INTERACTING PROTEIN 1 ( FTIP1 ) , is an essential regulator required for FT protein transport in Arabidopsis . Loss of function of FTIP1 exhibits late flowering under long days , which is partly due to the compromised FT movement to the shoot apex . FTIP1 and FT share similar mRNA expression patterns and subcellular localization , and they interact specifically in phloem companion cells . FTIP1 is required for FT export from companion cells to sieve elements , thus affecting FT transport through the phloem to the SAM . Our results provide a mechanistic understanding of florigen transport , demonstrating that FT moves in a regulated manner and that FTIP1 mediates FT transport to induce flowering .
The transition to flowering , which is crucial for the reproductive success , is the most dramatic phase change in flowering plants . Plants are able to adjust the timing of this transition in response to environmental conditions , such as photoperiod , temperature , and availability of nutrients . Classic experiments on the photoperiodic control of flowering in various plants have demonstrated that plant response to day length begins with the perception of photoperiod in leaves , followed by the transmission of a floral stimulus into the shoot apical meristem ( SAM ) , where flowers are generated instead of leaves . Such mobile floral stimulus moving from leaves to the SAM was proposed as “florigen” in the 1930s [1] . Since then , tremendous efforts have been made to understand the molecular nature of this signal . Recent findings have suggested that the proteins encoded by FLOWERING LOCUS T ( FT ) in Arabidopsis and its orthologs in other plant species are part of the long-sought florigen [2]–[6] . FT encodes a member of the phosphatidylethanolamine-binding protein family and acts as a crucial regulator that relays flowering signals from the photoperiod pathway to floral meristem identity genes in Arabidopsis , which is a long-day ( LD ) facultative plant [7]–[10] . Under LDs , FT mRNA expression is activated by the CONSTANS ( CO ) transcriptional regulator in the vascular tissues of leaves and displays circadian rhythm [8] , [11]–[14] . It has been suggested that long-distance movement of FT protein from leaves to the shoot apex through the phloem system plays a role in floral induction [2] , [4] , [5] . In the SAM , FT interacts with the bZIP transcription factor FLOWERING LOCUS D ( FD ) , which in turn activates the downstream floral meristem identity genes such as APETALA1 ( AP1 ) to initiate flower development [8] , [9] . Despite the remarkable progress in elucidating FT function , it is so far completely unknown whether and how FT protein transport is regulated . As the abundance of native FT protein is too low to be detectable , it has been hypothesized that simple diffusion of FT protein from companion cells to sieve elements might not be sufficient for transporting FT to the SAM [15] . Here we show that an endoplasmic reticulum ( ER ) membrane protein , FT-INTERACTING PROTEIN 1 ( FTIP1 ) , is required for FT protein transport in Arabidopsis . Loss of function of FTIP1 exhibits late flowering under LDs , which is partly due to the compromised FT movement to the SAM . FTIP1 and FT have similar mRNA expression patterns and subcellular localization , and they interact in vivo in phloem companion cells . Furthermore , FTIP1 is required for FT export from companion cells to sieve elements , thus affecting FT transport through the phloem to the SAM . Our results provide a mechanistic understanding of florigen transport and demonstrate that FT protein moves in a regulated manner and that FTIP1 is involved in mediating the export of FT protein from phloem companion cells to induce flowering .
To understand how FT function is regulated , we performed yeast two-hybrid screening to identify proteins that interact with FT . Approximately 3 million yeast transformants were screened and 66 colonies were identified on the selective medium ( Table S1 ) , among which a partial sequence belonging to an unknown protein with three C2 domains and one phosphoribosyltransferase C-terminal domain ( PRT_C ) was isolated ( Figure S1 ) . The corresponding gene ( At5g06850 ) was therefore named FT-INTERACTING PROTEIN 1 ( FTIP1 ) . We isolated two T-DNA insertional alleles , ftip1-1 ( Salk_013179 ) and ftip1-2 ( Salk_088086 ) , from Arabidopsis Biological Resource Center ( Figure 1A ) . The full-length FTIP1 transcript was undetectable in either homozygous mutant ( Figure 1B ) . Both ftip1-1 and ftip1-2 flowered late under LDs , but not under short days ( SDs ) ( Figure 1C , D; Table 1 ) , suggesting that FTIP1 plays a role in mediating the effect of photoperiod on flowering . We transformed ftip1-1 with a genomic construct ( gFTIP1 ) harboring a 5 . 1-kb FTIP1 genomic region including 2 . 1 kb of the upstream sequence , the 2 . 4-kb coding sequence , and 0 . 6 kb of the downstream sequence ( Figure S2A ) . Most ftip1-1 gFTIP1 T1 transformants exhibited similar or slightly late flowering time as compared to wild-type plants ( Figure 1E ) , demonstrating that FTIP1 is responsible for promoting flowering particularly under LDs . We tested FTIP1 expression in various tissues of wild-type plants using quantitative real-time PCR and found its highest expression in leaves and stems ( Figure S3 ) . To examine the detailed expression pattern of FTIP1 , we generated a FTIP1:β-glucuronidase ( GUS ) reporter construct in which the same 2 . 1-kb FTIP1 upstream sequence included in gFTIP1 for the gene complementation test was fused to the GUS reporter gene ( Figure S2A ) . We created 23 independent FTIP1:GUS lines , most of which showed similar GUS staining patterns . A representative line was selected to monitor the detailed expression pattern of FTIP1 . FTIP1:GUS showed specific GUS staining in vascular tissues of various plant organs ( Figure S2B–H ) . Notably , in developing seedlings during the floral transition occurring 7 d after germination , FTIP1:GUS and FT:GUS [14] shared similar GUS staining patterns in vascular tissues of cotyledons and rosette leaves , although the former had a relatively broad and intensive staining pattern ( Figure 2A ) . A cross-section of a primary leaf vein revealed that FTIP1:GUS expression was specifically located in the phloem including companion cells ( Figure 2B ) , which is similar to the FT:GUS expression pattern [14] . Neither FTIP1:GUS nor FT:GUS was expressed in the SAM ( Figure 2C , D; Figure S2C ) [14] . Furthermore , the late-flowering phenotype of ftip1-1 was rescued by the expression of FTIP1 coding sequence driven by the promoter of SUCROSE TRANSPORTER 2 ( SUC2 ) ( Figure 3A ) , which is active specifically in phloem companion cells [16] . These results suggest that FTIP1 functions in the phloem to promote flowering . Given that FTIP1 functions in flowering time control , we investigated whether its expression is regulated by known flowering genetic pathways . FTIP1 expression was not regulated by photoperiod and did not exhibit an obvious circadian rhythm under LDs ( Figure S4A , E ) . Similarly , vernalization treatment did not affect FTIP1 expression ( Figure S4B ) , and GA treatment did not affect FTIP1 expression and the flowering phenotype of ftip1-1 ( Figure S4C , D ) . In addition , FTIP1 expression was also not altered in several mutants tested in known flowering genetic pathways ( Figure S5 ) . These observations imply that flowering signals may not regulate FTIP1 function through affecting its mRNA levels . Next , we examined the subcellular localization of FTIP1 through monitoring the signal of the green fluorescent protein ( GFP ) fused with FTIP1 under the control of FTIP1 or SUC2 promoter , respectively . Both constructs could rescue the late flowering phenotype of ftip1-1 ( Figure 3A ) . However , we could not detect fluorescent signal from either SUC2:FTIP1:GFP ftip1-1 or FTIP1:FTIP1:GFP ftip1-1 transgenic lines , indicating that FTIP1 protein might be present at very low abundance in plant cells . Alternatively , we transiently expressed 35S:FTIP1:GFP with various fluorescent protein-tagged organelle markers in N . benthamiana leaf epidermal cells and found that FTIP1:GFP was mostly colocalized with an endoplasmic reticulum ( ER ) marker ( Figure 3B , C; Figure S6 ) [17] . We did not observe FTIP1:GFP signals in the nucleus ( Figure 3C ) . Notably , at the cell wall , FTIP1:GFP colocalized with callose deposition stained with aniline blue , which marks the position of plasmodesmata ( Figure 3C ) . To precisely localize FTIP1 , we performed immunoelectron microscopy on an FTIP1:4HA:FTIP1 ftip1-1 transgenic line , in which FTIP1:4HA:FTIP1 was able to rescue the flowering defect of ftip1-1 ( Figure 3A ) . The result revealed that 4HA:FTIP1 was specifically localized in phloem companion cells ( Figure 3D ) and plasmodesmata between companion cells and sieve elements ( Figure 3E , F; Figure S7 ) , where the ER membrane runs through . Several pieces of evidence , including the initial identification of FTIP1 as an FT interacting partner , similar tissue expression pattern of FTIP1 and FT , and similar late-flowering phenotype exhibited by ftip1 and ft mutants specifically under long days , point to a possible role of FTIP1 in mediating FT function in the control of flowering time . Thus , we further carried out a detailed analysis of the interaction between FTIP1 and FT . As revealed in our yeast two-hybrid screening , a truncated FTIP1 protein devoid of the PRT_C domain interacted with FT in both yeast two-hybrid and GST pull-down assays ( Figure 4A–C ) , whereas no interaction was detected using the full-length FTIP1 ( unpublished data ) . Since the PRT_C domain of FTIP1 was predicted to be a membrane-targeted domain according to a protein topology analysis ( Figure S1 ) , the full-length FTIP1 protein might not be in the membrane-bound state in yeast cells or under in vitro conditions and thus might undergo inappropriate folding , which prevents its interaction with FT . Alternatively , in yeast two-hybrid assay the full-length FTIP1 protein might be membrane-bound and unable to reconstitute a functional transcription factor in the yeast nucleus to drive the reporter gene expression . We transiently expressed 35S:FTIP1:GFP with 35S:FT:RFP in N . benthamiana leaf epidermal cells and revealed that both FTIP1:GFP and FT:RFP were colocalized to ER connected to the nuclear envelope ( Figure S8 ) . However , in contrast to FTIP1:GFP , FT:RFP was also localized in the nucleus , which is consistent with a previous observation [9] . These results indicate that FTIP1 may not directly mediate FT function in transcriptional regulation of other target genes . To test whether and how FT interacts with FTIP1 in vivo , we performed in situ Proximity Ligation Assay ( PLA ) [18] , in which dual recognition of target proteins by pairs of affinity probes generates an amplifiable DNA reporter molecule that serves as a surrogate marker for interacting proteins , to examine the subcellular localization of FT and FTIP1 interaction at single-molecule resolution in the leaves of 11-d-old SUC2:FT:GFP; FTIP1:4HA:FTIP1 transgenic plants . PLA signals visualized as small red dots were specifically detected in the phloem companion cells of SUC2:FT:GFP; FTIP1:4HA:FTIP1 , but barely in those transgenic plants containing only SUC2:FT:GFP , FTIP1:4HA:FTIP1 , or SUC2:GFP; FTIP1:4HA:FTIP1 ( Figure 4D , E ) . This result demonstrates that FT and FTIP1 physically interact in close proximity in phloem companion cells . The findings on the interaction between FT and FTIP1 , and FTIP1 localization to ER and plasmodesmata prompted us to hypothesize that FTIP1 may regulate FT export from phloem companion cells . To this end , we first examined whether FTIP1 affects FT transport to the SAM during the floral transition . We generated a SUC2:FT:GFP transgenic line as previously described [2] . As this transgenic allele could significantly rescue the late-flowering phenotype of the FT null mutant , ft-10 ( Table 1 ) , we further crossed this SUC2:FT:GFP allele with ftip1-1 and 35S:FTIP1 . Confocal analysis of the distribution of FT:GFP fusion protein revealed that in 11-d-old seedlings , which were undergoing the floral transition , FT:GFP was clearly detected in the inner cone-like region of the SAM in wild-type background , but not in ftip1-1 ( Figure 5A ) . In contrast , the distribution of free GFP protein was comparably undetectable in the inner region of the SAM in wild-type and ftip1-1 ( Figure S9A ) , indicating a specific effect of FTIP1 on FT:GFP distribution in the SAM during the floral transition . In agreement with the above observations , SUC2:FT:GFP ftip1-1 flowered later than SUC2:FT:GFP ( Table 1 ) . Since the abundance of FT:GFP mRNA and protein in SUC2:FT:GFP was not downregulated in ftip1-1 ( Figure 6A–C ) , the difference in FT:GFP distribution in the SAM between wild-type and ftip1-1 plants suggests a role of FTIP1 in regulating FT transport rather than FT mRNA or protein abundance . As FTIP1 was expressed in the phloem ( Figure 2 ) and its protein was localized in phloem companion cells ( Figure 3D–F; Figure 4D ) , we examined whether FTIP1 affects FT transport from companion cells to sieve elements in a newly created SUC2:FT:9myc line in wild-type and ftip1-1 backgrounds using immunoelectron microscopy ( Figure 5B ) . This SUC2:FT:9myc transgenic allele substantially rescued the late-flowering phenotype of ft-10 ( Table 1 ) , indicating that FT:9myc retains the biological function of endogenous FT protein . Signals corresponding to FT:9myc could be specifically detected by anti-myc antibody in the phloem of the transgenic plants harboring SUC2:FT:9myc ( Figure 5B; Figure S10 ) . Quantitative analysis of labeling density of FT:9myc in companion cell-sieve element complexes showed that although all sections from SUC2:FT:9myc and SUC2:FT:9myc ftip1-1 displayed FT:9myc labeling in companion cells ( Figure 5B , lower right panel ) , there was an approximate 3-fold enrichment of labeling density in ftip1-1 over wild-type background ( Figure 5B , lower left panel ) . More importantly , we detected FT:9myc labeling in sieve elements in nearly 80% of the wild-type sections , whereas only 4% of ftip1-1 sections displayed FT:9myc labeling in sieve elements ( Figure 5B , lower right panel ) . In addition , the labeling density of FT:9myc in sieve elements was much higher in wild-type than in ftip1-1 ( Figure 5B , lower left panel ) . Thus , in the absence of FTIP1 , FT:9myc accumulated in companion cells and its transport to sieve elements was compromised . In agreement with this result , SUC2:FT:9myc ftip1-1 displayed later flowering than SUC2:FT:9myc ( Table 1 ) . As ftip1-1 also delayed flowering in SUC2:FT and SUC2:GFP:CO where CO directly promotes the endogenous FT expression ( Table 1 ) [12] , it seems that FTIP1 similarly affects the promotive effect of untagged FT protein on flowering as other FT fusion proteins used in this study . These observations support that FTIP1 regulates FT export from phloem companion cells to sieve elements , thus affecting FT transport through the phloem to the SAM . Consistent with this conclusion , the early-flowering phenotype caused by expression of FT or FT:GFP under the control of the KNAT1 promoter , which is active in the SAM [12] , was not affected by ftip1-1 ( Table 1 ) . Unlike other flowering promoters , overexpression of FTIP1 surprisingly caused late flowering ( Figure S11; Table 1 ) . Confocal analysis showed that the expression of FT:GFP protein in the inner region of the SAMs of 11-d-old seedlings was substantially lower in 35S:FTIP1 than in wild-type plants ( Figure 5A; Figure S9A ) . In the primary leaf vein , FT:GFP driven by the SUC2 promoter was exclusively detected in phloem companion cells in wild-type background , whereas in 35S:FTIP1 , the distribution of FT:GFP signals was detected in both phloem companion cells and xylem parenchyma cells ( Figure 5C ) . However , the free GFP driven by the SUC2 promoter remained in phloem companion cells of 35S:FTIP1 as compared to wild-type plants ( Figure S9B ) . These observations demonstrate that that ectopic expression of FTIP1 specifically deregulates the transport of FT:GFP protein out of the phloem system , an effect previously shown for a viral movement protein [19] . This could compromise the eventual distribution of FT:GFP in the SAM of 35S:FTIP1 and thus delay flowering . During the floral transition , FT interacts with FD in the SAM to promote the expression of AP1 and other flowering genes such as SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 ( SOC1 ) [8] , [9] , [20] . As expected , the expression of these genes was downregulated in ftip1-1 in which FT transport is defective ( Figure S12A ) . Surprisingly , FT expression was also downregulated in ftip1-1 , whereas the expression of CO , a direct upstream activator of FT , was not significantly changed ( Figure S12 ) . As FTIP1 protein is not localized in the nucleus , it is unlikely that FTIP1 directly controls FT transcription . To address whether FTIP1 could regulate the stability of FT transcripts , we compared the levels of FT transcripts generated from the native and SUC2:FT:GFP transgenic loci . Although steady-state levels of native FT expression were downregulated in ftip1-1 , total FT expression including native FT and transgenic FT:GFP expression remained unchanged in SUC2:FT:GFP ftip1-1 ( Figure 6A , B ) . In addition , the abundance of FT:GFP fusion protein remained unchanged in wild-type and ftip1-1 backgrounds ( Figure 6C ) . These results suggest that FTIP1 may not be directly involved in regulating FT mRNA or protein stability . Meanwhile , we observed downregulation of native FT expression in SUC2:FT:GFP ( Figure 6A , B ) and reduced FT:GUS staining in SUC2:FT ( Figure 6D ) . These results are in agreement with the observation in a previous study [2] implying that an excessive accumulation of FT protein in phloem companion cells caused by the SUC2 promoter might directly or indirectly result in a reduction in native FT mRNA expression through a negative feedback loop . This may explain the observed downregulation of native FT expression in ftip1-1 , where defective export of FT protein causes accumulation of FT protein in phloem companion cells ( Figure 5B ) .
Our results have demonstrated that FTIP1 and FT share similar mRNA expression patterns and subcellular localization , and that they interact in vivo in phloem companion cells . During the floral transition , the FT:GFP accumulation at the SAM is compromised in ftip1 mutants , which eventually exhibit late flowering under LDs . Consistently , FTIP1 is required for FT:9myc export from phloem companion cells to sieve elements , thus affecting the flowering time of SUC2:FT:9myc . In addition , overexpression of FTIP1 causes the transport of FT:GFP out of the phloem system , which also results in late flowering . These observations suggest that FT protein moves from phloem companion cells to sieve elements in a regulated manner and that a subtle regulation of FTIP1 activity is indispensable for the export of FT protein from phloem companion cells to induce flowering . We envisage that in addition to FTIP1 and FT , florigen transport should involve other relevant regulators . First , although the transport of FT:9myc protein from phloem companion cells to sieve elements in ftip1-1 is significantly compromised , it is not completely abolished . This implies either that there is a basal level of diffusion of FT protein or that FT transport also depends on other regulators that share a redundant function with FTIP1 in mediating FT export from phloem companion cells . Furthermore , previous examinations of the spatial distribution of FT:GFP fusion protein in both Arabidopsis and rice have shown that FT:GFP accumulates in the rib zone beneath the SAM in a conical shape [2] , [3] , indicating that the movement of FT protein from phloem to the rib zone is not a simple diffusion process . As FTIP1 is clearly not expressed in the whole SAM ( Figure 2C , D ) , regulation of FT protein transport from the phloem stream to the rib zone might also involve other regulators . The requirement of other regulators for FT protein transport is supported by the genetic analysis showing that ft mutants display much later flowering than ftip1 ( Table 1 ) . Potential candidates for these regulators include FTIP1 homologs ( Figure S13A ) because some combinations of ftip1 with loss-of-function mutants of FTIP1 homologs show much later flowering than any single mutant ( unpublished data ) . Second , the late-flowering phenotype of ft mutants is further enhanced by ftip1-1 ( Table 1 ) , indicating that FTIP1 may be required for transporting other flowering molecules in addition to FT . A potential candidate could be TWIN SISTER OF FT ( TSF ) , which encodes another phosphatidylethanolamine-binding protein with very high sequence similarity with FT [21] , [22] . Mutation of TSF further enhances the late flowering of ft mutants , and the resulting double mutants fail to accelerate flowering in response to LD conditions [21] , [22] . The expression domain of TSF also overlaps with that of FTIP1 [21] . Furthermore , loss of function of FTIP1 does not further delay flowering of ft-10 tsf-1 under LD conditions ( Table 1 ) . These data support that TSF functions redundantly with FT and could be another molecule whose transport is affected by FTIP1 . As both FTIP1 and FT proteins are localized to ER , regulation of FT movement by FTIP1 across the border between companion cells and sieve elements might be partly mediated by a continuous ER network within plasmodesmata [23] , [24] . In plasmodesmata , the ER becomes appressed to form the central axial desmotubule surrounded by the plasma membrane continuum between adjacent cells [25] . Although it has been suggested that the desmotubule is not the main route for plasmodesmatal transport , some molecules are known to be transported through this channel [26] . In contrast , the space between the desmotubule and the plasma membrane , which is referred as the cytoplasmic sleeve , is the proposed place where the general trafficking of molecules and ions occurs [25] . Because FTIP1 possesses a membrane-targeted PRT_C domain ( Figure S1 ) and is localized to plasmodesmata ( Figure 3C , E , F ) , it is likely that the C-terminus of FTIP1 is anchored to the desmotubule . How FTIP1 is oriented in plasmodesmata is an important question as its N-terminus , which is included in the region that interacts with FT protein ( Figure 4A–C ) , might face either the cytoplasmic sleeve or the interior of the desmotubule . Further addressing this question will help to identify the route where FT protein passes through plasmodesmata and other possible factors involved in FT transport . Based on the size of FT:GFP , the route through the cytoplasmic sleeve might be possible for FT transport as the current understanding is that molecules larger than 27 kDa do not move readily through desmotubule [23] . The presence of C2 domains and a transmembrane domain in FTIP1 and its close homologs in Arabidopsis makes them topologically resemble synaptotagmins ( Figure S13A ) that constitute a family of membrane-trafficking proteins widely found in plants and animals . In Arabidopsis , the synaptotagmin SYTA has been shown to regulate endosome recycling and movement protein-mediated trafficking of plant virus genomes through plasmodesmata [27] . Our finding on the function of FTIP1 in mediating FT export from phloem companion cells to sieve elements , together with the proposed SYTA function , implies that synaptotagmin-like proteins may serve as essential regulators that mediate the transport of macromolecules in plants . Another FTIP1-like gene , QUIRKY ( QKY; At1g74720 ) , has been suggested to contribute to plant organ organogenesis mediated by the receptor-like kinase STRUBBELIG [28] , implying a role of QKY in intercellular signaling . As FTIP1-like proteins are highly conserved in the angiosperms ( Figure S13B ) , further investigation of FTIP1 and its homologs might shed light on the conserved mechanisms underlying which flowering plants regulate cell-to-cell communication to coordinate the growth and development .
Arabidopsis plants were grown at 22°C under long days ( 16 h light/8 h dark ) or short days ( 8 h light/16 h dark ) . The mutants ftip1-1 , ftip1-2 , co-1 , gi-1 , ft-1 ( Ler ft-1 introgressed into Col ) , ft-10 , tsf-1 , soc1-2 , agl24-1 , fve-3 , and svp-41 are in Columbia ( Col ) background , while co-2 , fca-1 , fpa-1 , and fve-1 are in Landsberg erecta ( Ler ) background . To construct 35S:FTIP1 , the cDNA encoding FTIP1 was amplified with primers and cloned into pGreen-35S [29] . For the complementation test , a 5 . 1-kb FTIP1 genomic fragment ( gFTIP1 ) was amplified and cloned into pHY105 [28] . Based on this construct , FTIP1:FTIP1:GFP and FTIP1:4HA:FTIP1 were generated using a modified QuikChange Site-Directed Mutagenesis approach [30] . The cDNAs encoding GFP and 4HA were amplified . The resulting PCR fragments were annealed to the methylated template plasmid DNA containing gFTIP1 and elongated with the PfuTurbo DNA polymerase ( Stratagene ) . Upon DpnI digestion , the mutated plasmids containing either GFP or 4HA were recovered from E . coli transformation . To construct pGreen-SUC2 and pGreen-KNAT1 , SUC2 and KNAT1 promoters were amplified from Col genomic DNA and cloned into pHY105 [28] . To construct SUC2:FTIP1 , the cDNA encoding FTIP1 was amplified and cloned into pGreen-SUC2 . Based on SUC2:FTIP1 and 35S:FTIP1 , SUC2:FTIP1:GFP and 35S:FTIP1:GFP were generated using the same modified QuikChange Site-Directed Mutagenesis approach [30] for creating FTIP1:FTIP1:GFP . To construct SUC2:FT and KNAT1:FT , the cDNA encoding FT was amplified and cloned into pGreen-SUC2 and pGreen-KNAT1 , respectively . Based on the constructs of SUC2:FT and KNAT1:FT , SUC2:FT:GFP , SUC2:FT:9myc and KNAT1:FT:GFP were generated using the same modified QuikChange Site-Directed Mutagenesis approach [30] for creating FTIP1:FTIP1:GFP . To construct 35S:FT:RFP , the cDNA encoding RFP was amplified and cloned into pGreen-35S to generate pGreen-35S-RFP . The cDNA encoding FT was subsequently amplified and cloned into pGreen-35S-RFP . To construct FTIP1:GUS , the 2 . 1-kb FTIP1 5′ upstream sequence was amplified and cloned into pHY107 [29] . All transgenic plants were created using the floral dip method [31] and screened by Basta on soil . All vectors used in yeast two-hybrid assays were from Clontech . The coding sequence of FT was cloned into pGBKT7 to produce BD-FT , which was used as a bait to screen cDNA library ( CD4-30 from ABRC ) for identifying interacting proteins of FT . Selection was performed on medium lacking histidine , tryptophan , and leucine ( SD-His/-Trp/-Leu ) supplemented with 30 mM 3-amino-1 , 2 , 4-triazole . To verify the interaction between FT and FTIP1 , various versions of FTIP1 coding sequences were cloned into pGADT7 . The resulting vectors were co-transformed with BD-FT , and the transformed cells were selected on SD-His/-Trp/-Leu medium supplemented with 30 mM 3-amino-1 , 2 , 4-triazole . β-galactosidase assays were performed according to the Yeast Protocols Handbook ( Clontech ) . The cDNA encoding FT was cloned into the pGEX-4T-1 vector ( Pharmacia ) and introduced into E . coli Rosetta ( DE3 ) ( Novagen ) . Transformed cells were cultured until the OD600 nm reached 0 . 6 , and IPTG was added to a final concentration of 0 . 6 mM to start induction . After overnight induction at 16°C , cells were collected and homogenized with lysis buffer ( 10 mM Tris-HCl pH 8 . 0 , 150 mM NaCl , 1 mM EDTA , 1% Triton-100 , and 10 mM PMSF ) . The soluble GST fusion proteins were extracted and immobilized on glutathione sepharose beads ( Amersham Biosciences ) for subsequent pull-down assays . The FTIP1 N-terminal fragment containing the three C2 domains ( N501 ) was cloned into pGBKT7 vector ( Clontech ) . The resulting plasmid was added to the TNT T7 Quick Coupled Transcription/Translation Systems ( Promega ) to synthesize myc-FTIP1 ( N501 ) protein . The resulting fusion protein was then incubated with the immobilized GST and GST fusion proteins . Proteins retained on the beads were resolved by SDS-PAGE and detected with anti-myc antibody ( Santa Cruz Biotechnology ) . The overnight Agrobacterium cultures with a desired expression vector ( 35S:FTIP1:GFP , various RFP- or CFP-tagged organelle markers , 35S:FT:RFP , or 35S:GFP ) were harvested and resuspended with infiltration buffer ( 10 mM MES pH 5 . 6 , 10 mM MgCl2 , and 100 µM acetosyringone ) with OD600 nm at 0 . 4 . To compare protein localization , equal volumes of infiltration solutions containing different expression vectors were mixed together and incubated for 3 h at room temperature . Infiltration solutions were infiltrated into the abaxial surface of 3-wk-old tobacco ( Nicotiana benthamiana ) leaves with syringes . The leaves were examined 2 d after infiltration under a confocal microscope . Tissues were infiltrated with staining solution ( 50 mM sodium phosphate buffer , pH 7 . 0 , 0 . 5 mM potassium ferrocyanide , 0 . 5 mM potassium ferricyanide , and 0 . 5 mg/mL X-Gluc ) in a vacuum chamber , and subsequently incubated with staining solution at 37°C overnight . For sectioning , samples were dehydrated through an ethanol series , an ethanol/histoclear series , and finally embedded in paraplast ( Sigma ) . Samples were then orientated and sectioned at a thickness of 3 µm with a microtome . Total RNA was isolated with RNeasy Plant Mini Kit ( Qiagen ) and reverse-transcribed with ThermoScript RT-PCR System ( Invitrogen ) according to the manufacturers' instructions . Real-time PCR was performed in triplicates on 7900HT Fast Real-Time PCR system ( Applied Biosystems ) with SYBR Green PCR Master Mix ( Applied Biosystems ) . The difference between the cycle threshold ( Ct ) of the target gene and the Ct of TUB2 ( ΔCt = Cttarget gene−Cttubulin ) was used to obtain the normalized expression of target genes , which corresponded to 2−ΔCt . Expression analysis was performed with at least three biological replicates . Primers for real-time PCR are listed in Table S2 . Plant tissues were collected and fixed with ice-cold 4% paraformaldehyde ( PFA; Sigma-Aldrich ) at pH 7 . 0 in a vacuum chamber . A serial PFA/sucrose change was applied till the tissues were finally equilibrated in PFA with 20% sucrose . Tissues were then embedded in 1 . 5% agarose gel , placed onto the microtome tissue holder , and flash frozen with liquid nitrogen . Tissues were cut in cryo-microtome with 20 µm thickness , and the sections were placed on slides . After complete drying , the slides were rehydrated with 100 mM Tris pH 8 , 50 mM EDTA , and permeabilized with proteinase K ( 1 µg/ml ) in the same buffer for 10 min at room temperature . Slides were washed with 2 mg/ml glycine followed by washing with PBS solution . Chlorophyll molecules were subsequently removed by incubating the slides with 1∶1 acetone/methanol mixture twice for 5 min . After drying , slides were rehydrated with PBS and finally treated with 4% paraformaldehyde for 10 min followed by washing with PBS solution . In situ Proximity Ligation Assay ( PLA ) was performed with Duolink kit ( Olink Bioscience ) with minor modifications . The above treated slides were firstly blocked with 2% Bovine Serum Albumin in 100 mM Tris pH 7 . 5 , 150 mM NaCl , and 0 . 3% Triton X-100 for 45 min at 37°C , and probed with the mixture of anti-GFP and anti-HA antibodies diluted in the blocking solution ( 1∶60 ) for 45 min at 37°C . The slides were washed three times and probed with diluted PLUS and MINUS PLA probes for 1 h at 37°C and subsequently washed 5 times . The slides were further incubated with the ligation solution , washed , and subsequently incubated with the amplification-polymerase solution with all components provided in the kit . After signal amplification , the slides were washed and mounted with PBS solution for further observation . Samples were fixed with paraformaldehyde-glutaraldehyde solution ( 2% and 2 . 5% , respectively ) and imbedded with LR white medium ( EMS ) . Ultra-thin sections ( 85 nm ) were cut and mounted on nickel grids . The grids were blocked with 1% BSA in TTBS ( 20 mM Tris , 500 mM NaCl , and 0 . 05% Tween-20 , pH 7 . 5 ) for 30 min and subsequently incubated with anti-HA or anti-myc antibody at 1∶5 ( v/v ) for 1 h at room temperature . The grids were washed with TTBS for three times and further incubated with 15 nm gold-conjugated goat anti-mouse antibody ( EMS ) that was diluted 1∶20 with blocking solution . After 40 min of incubation , the grids were washed with TTBS for three times and with distilled water twice . Tissue staining was performed with 2% uranyl acetate for 15 min at room temperature , and pictures were taken by transmission electron microscope ( Jeol JEM-1230 ) . For quantitative analysis of immunogold labeling , micrographs of randomly photographed immunogold-labeled transverse sections of the first rosette leaves of 15-d-old seedlings with various genetic backgrounds were measured as previously reported [32] . The data were presented as the mean number of gold particles per µm2 plus or minus standard deviation . The projected cell area was measured by a LI-3100C area meter ( Li-Cor ) . We analyzed 56 individual sections from eight different leaves of each genotype for calculating the density of gold particles over the projected cell area . Statistical analysis was performed using a two-tailed unpaired Student's t test . Two-tailed test results were considered statistically significant at p<0 . 05 .
|
The transition to flowering is the most dramatic phase change in flowering plants and is crucial for reproductive success . Such a transition from vegetative to reproductive growth is controlled by seasonal changes in day length . Studies originally performed in the 1930s were the first to suggest that day length is perceived by a plant's leaves; by contrast , flower formation takes place in the shoot apical meristem ( the tip of the shoot that gives rise to plant organs , such as leaves and flowers ) . The term “florigen” was later proposed to describe a mobile floral stimulus that moves from leaves to the shoot apical meristem to induce flowering . It is only recently that FLOWERING LOCUS T ( FT ) in Arabidopsis , and its orthologs in various other plant species , was identified as being florigen , but how florigen is transported in plants remains completely unknown . Here , we report that a novel ER membrane protein , FT-INTERACTING PROTEIN 1 ( FTIP1 ) , interacts with FT in companion cells of the phloem ( a specialized type of parenchyma cell in the phloem of the plant's vascular system ) and mediates FT protein movement from companion cells to sieve elements ( the conducting cells of the phloem ) , thus affecting FT transport to the shoot apical meristem in Arabidopsis . To our knowledge , this study reveals the first regulator that is required for florigen transport and offers new insights into possible florigen transport mechanisms in other flowering plants .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"plant",
"science",
"plant",
"growth",
"and",
"development",
"plant",
"biology",
"biology"
] |
2012
|
FTIP1 Is an Essential Regulator Required for Florigen Transport
|
Although L1 sequences are present in the genomes of all placental mammals and marsupials examined to date , their activity was lost in the megabat family , Pteropodidae , ∼24 million years ago . To examine the characteristics of L1s prior to their extinction , we analyzed the evolutionary history of L1s in the genome of a megabat , Pteropus vampyrus , and found a pattern of periodic L1 expansion and quiescence . In contrast to the well-characterized L1s in human and mouse , megabat genomes have accommodated two or more simultaneously active L1 families throughout their evolutionary history , and major peaks of L1 deposition into the genome always involved multiple families . We compared the consensus sequences of the two major megabat L1 families at the time of their extinction to consensus L1s of a variety of mammalian species . Megabat L1s are comparable to the other mammalian L1s in terms of adenosine content and conserved amino acids in the open reading frames ( ORFs ) . However , the intergenic region ( IGR ) of the reconstructed element from the more active family is dramatically longer than the IGR of well-characterized human and mouse L1s . We synthesized the reconstructed element from this L1 family and tested the ability of its components to support retrotransposition in a tissue culture assay . Both ORFs are capable of supporting retrotransposition , while the IGR is inhibitory to retrotransposition , especially when combined with either of the reconstructed ORFs . We dissected the inhibitory effect of the IGR by testing truncated and shuffled versions and found that length is a key factor , but not the only one affecting inhibition of retrotransposition . Although the IGR is inhibitory to retrotransposition , this inhibition does not account for the extinction of L1s in megabats . Overall , the evolution of the L1 sequence or the quiescence of L1 is unlikely the reason of L1 extinction .
L1 ( LINE-1 , Long INterspersed Element-1 ) belongs to the superfamily of autonomously replicating , retrotransposable elements that lack long terminal repeats . Functional L1s are 6 , 000–7 , 000 bp long and made up of a 5′ untranslated region ( 5′UTR ) , two non-overlapping open reading frames ( ORFs ) known as ORF1 and ORF2 , an intergenic region ( IGR ) usually less than 100 bp and a 3′UTR followed by a poly-adenosine sequence [1] . The proteins encoded by both ORFs are strictly required for L1 retrotransposition and have very strong cis-preference [2] , [3] . The function of the IGR is less well characterized , but it is known to be indispensable for the translation of human ORF2 protein [4] and to serve as an internal ribosome entry site ( IRES ) in mice [5] . There is considerable evidence that transposable elements , including L1s , have significant effects on the genome . L1 retrotransposition is one of the major sources of mutagenesis and genome instability [6] , [7] . Besides their copy-and-paste retrotransposition mechanism that interrupts genes and disrupts the normal splicing of messenger RNAs [8] , L1s also cleave genomic DNA with the endonuclease they encode [9]–[13] and are sites of ectopic recombination due to their homology to each other and prevalence throughout the genome [14]–[18] . L1s and their dependents may be occasionally co-opted to provide host functions . For example , they may serve as the source of new genes [8] or structural chromosome components [19] , or regulate genes in their vicinity by various mechanisms [20]–[22] . They have also been proposed to play a role in X chromosome inactivation [23]–[25] , neuro-plasticity [26]–[28] and regulatory functions [29] . L1s have been coevolving with their mammalian host genomes since before the eutherians and metatherians diverged [30] more than 160 million years ago ( MYA ) [31] . The tempo of L1 retrotransposition can vary both between species and at different time intervals within species [32]–[35] . They evolve as master lineages such that closely related active L1 copies succeed the older masters and become new major contributors to the total retrotransposition events [33] , [36]–[38] . Most species are dominated for long periods of time by a single such master lineage [1] , although multiple lineages are occasionally active at the same time [32] , [35] , [39] . Retrotransposition of the L1 population is extremely inefficient and few new active elements are produced , with the vast majority of new inserts being 5′ truncated pseudogenes . There are over 500 , 000 copies of L1 in the human reference genome [40] , but only 80–100 of the L1s in an average human genome are estimated to be full-length and retrotranspositionally competent , with just six of these contributing more than 80% of the total L1 activity . These six elements are closely related; all belong to the youngest family of human L1s , and four of them belong to the youngest clade within that family [41] . Because there is no known mechanism for precise excision of L1s from the genome , old elements accumulate and make up 15–20% of a typical mammalian genome [40] , [42] . These ‘fossil’ sequences make it possible to track the activity of L1s within a particular mammalian clade back many millions of years . One possible reason for this unusual pattern of L1 evolution is that L1s are epigenetically silenced [43] , [44] and highly regulated by a set of host defense mechanisms [45]–[48] , especially in germline cells . Given the strong host defenses controlling L1 activity , it might seem reasonable to expect L1 extinctions among mammalian lineages . To clarify the terms related to loss of L1 activity in this work , we refer to a period of low L1 activity as “quiescence” and complete loss of L1 activity as “extinction . ” Indeed , quiescence or extinction of L1 has been proposed several times in the literature [32] , [49]–[54] , but few of these cases have been examined in a phylogenetic context to convincingly demonstrate that extinction , and not simply quiescence , best explains the lack of recent L1 insertions into the genome . Because L1s are transmitted vertically with no evidence of horizontal transmission among mammals , ancient L1 extinctions would affect all subsequent species and should be the most easily identified and confirmed . One well-documented case of L1 extinction occurred in the ancestor of the megabat family , Pteropodidae , which is the focus of this study . The L1 extinction was verified in 11 sampled genera within Pteropodidae , but did not affect other families of bats . The ancestor of the megabats had two active L1 lineages , both of which became extinct at about the same time at least 24 MYA [50] . In this study , the evolutionary history of L1s prior to their extinction in megabats was explored by data-mining the unassembled genome of Pteropus vampyrus , the first publicly available genome trace files of the megabat family . At the time of L1 extinction , P . vampyrus contained two active L1 lineages . We determined that these lineages likely diverged before the origin of bats . We reconstructed the master element of the more active lineage at the time of L1 extinction and compared its structure to other active L1s , noting particularly that the IGR between the two ORFs is dramatically longer than that of the well-characterized L1s of human and mouse . Finally , we created chimeric L1s between the reconstructed megabat L1 and a human L1 to test the ability of the extinct megabat L1 to support retrotransposition in tissue culture and we manipulated the IGR to explore its effect on retrotransposition .
To investigate the history of L1 retrotransposition in the megabats , we identified subfamilies using COSEG in RepeatMasker [55] based on shared , co-segregating sites within 575 bp of the 3′ end of ORF2 . These were designated subfamilies 0–63 using the convention of the program . The consensus sequences of these subfamilies were subjected to phylogenetic analysis and the phylogenetic relationships were used to identify families with the stipulation that the pairwise distances between subfamilies within a family be no greater than 3 . 5% . This distance was based on the observed phylogenetic clustering of subfamily consensus sequences . Given that the L1 masters are constantly being replaced during evolution , perfect designation within large families is not possible . The 3 . 5% threshold was chosen according to practical observations to cluster closely related subfamilies without inflating the number of families . This method identified 16 L1 families that account for the peaks of L1 fixation in the megabat genome ( Figure 1 and Table S1 ) . Previous work indicated that two major lineages of L1 were active at the time of L1 extinction in megabats [50] . Full-length consensus sequences from two time points in the evolution of each lineage can be found in RepBase [56] , [57] , designated L1-1_PVa to L1-4_PVa . COSEG analysis confirms and extends this history . Lineage 1 corresponds to families 1A ( L1-2_PVa ) , 1B ( L1-3_PVa ) and 1C . Lineage 2 corresponds to families 2A ( L1-1_PVa ) , 2B ( L1-4_PVa ) , 2C and 2D . It is clear that these two lineages existed prior to the emergence of the bats since families 2C and 2D are not bat-specific , but are closely related to elements found in various Laurasiatheria species . The older L1 families identified in our work ( 5–11 ) have high identity to the L1 families shared by all placental mammals [58] and by the Laurasiatheria superorder [59] . Smit et al . [58] designated the ancestral mammalian L1 families from most recent to oldest as L1MA , L1MB , L1MC , L1MD and L1ME . Subfamilies within each family are identified by number , with 1 being the most recent . The bottom panel of Figure 1 places megabat L1 dynamics in the context of these ancestral L1 families and the extant L1 lineages of primates and rodents . The relationship between the COSEG subfamilies , families and the ancestral L1s are summarized in Table S1 . To examine the activity and extinction of L1s in megabats , we extracted 79 , 978 L1 sequences from the ORF2 of L1s in the ∼2× unassembled shotgun sequence of the P . vampyrus genome ( Baylor College of Medicine ) and assigned them to one of the subfamilies described above based on sequence similarity . The age of each sequence was approximated by its percent identity to the subfamily consensus – the higher the percent identity , the younger the sequence . Subfamilies were combined into their designated families as determined by phylogenetic analysis ( described above ) and the age distribution was determined for each family . Taking all families together , we observed periodic fluctuations in the number of L1s fixed in the genome ( Figure 1 , top ) . At least two large waves of L1 fixation in megabats can be identified in the lineages described above with peaks at 92–93 . 5% and 87 . 5–89% similarity to subfamily consensus sequences ( Figure 2 ) . Each peak corresponds to activity of two or more families and to multiple lineages . The most recent peak , accounting for 25% of the L1s detected in the megabat genome , corresponds to families 1A and 2A and is megabat-specific . No more recent waves of retrotransposition can be identified , consistent with the extinction of L1 retrotransposition in the common ancestor of megabats ∼24 MYA [50] . The next peak , accounting for 13% of detected L1s , corresponds to activity in families 1B , 2B and 2C . A third peak , accounting for 12% of detected L1s , resides at 84 . 5–85 . 5% and corresponds to families 2D and 3; this peak likely represents retrotransposition prior to the origin of bats . Older waves of L1 fixation are also evident and correspond to ancestral mammalian L1 families . The dynamics of families within lineages 1 and 2 are not perfectly consistent with short bursts of retrotransposition followed by long periods of quiescence . Given the evolutionary pattern of L1 as master lineages , most L1 sequences evolve neutrally after their insertion into the genome . Therefore , the distribution of mutations in elements inserted at the same time should follow a Poisson distribution ( i . e . , the mean divergence from the consensus is expected to be equal to the variance of the distribution ) . However , the mean of each family is 1–2% larger than the peak , indicating that the variance of the distribution is higher than that of a Poisson distribution . This increased variance could be due to sequence differences between active L1s in the same subfamily at the time of transposition , a wave of retrotransposition over an extended period of time , errors introduced during L1 retrotransposition , technical problems with the analysis , or some combination of these . Technical issues might include false detection by RepeatMasker , incorrect assignment of some elements to their lineage or combining small lineages with larger ones , for example . Interestingly , the highest copy number peak is for family 1A , one of the two youngest detectable lineages active just prior to L1 extinction . This peak accounts for 18% of the total L1s detected in the megabat genome . We sought to reconstruct a full-length version ( minus the UTRs , which are difficult to accurately reconstruct ) of the more active L1 lineage in megabats at the time of L1 extinction , synthesize it and test its activity in a tissue culture assay . It was not possible to reconstruct the less active lineage with confidence because the copy number , especially in the 5′ end , is too low . Since the extinction of megabat L1 retrotransposition happened in the common ancestor of the family , the retrotransposition history of L1 in P . vampyrus represents that of the whole Pteropodidae family . Reconstruction was conducted on the P . vampyrus genome using a consensus-based method , with curated correction of CpG sites . We performed this reconstruction independently , without reference to RepBase [56] , [57] , thus the RepBase reconstruction served as a way to assess the quality of our reconstruction and a benchmark for problematic areas . Our reconstructed megabat L1 ( GenBank accession number KF796623 ) has 99 . 7% identity to the RepBase reconstruction ( RepBase Reports 10: ( 3 ) , 474-474 , 2010 , available at http://www . girinst . org/2010/vol10/issue3/L1-2_PVa . html ) at the nucleotide level , with six differences ( two in ORF1 and four in ORF2 ) at the amino acid level . The amino acid differences were examined individually in the original alignments: three resulted from ambiguous nucleotides or frame shifts in the RepBase reconstruction , one from CpG site correction and two from variable sites which we called differently than RepBase . None of these differences were at sites of conserved amino acids ( see below ) . Note that although RepBase designation L1-2_PVa suggests that this sequence falls within lineage 2 , we follow the precedence of Cantrell et al . [50] to designate it as a member of lineage 1 . We compared the reconstructed L1 to the most recently active consensus sequences from 31 diverse mammalian species ( Table S2 and Text S1 and S2 ) . Sequences are taken from RepBase except five which we reconstructed from trace files , including a rodent species carrying dead L1s , Oryzomys palustris . As noted in the Materials and Methods , several sequences were edited to restore ORFs . These alterations were generally within A-rich tracts , which are common in L1s and difficult to reconstruct with confidence . Since the 5′ end of ORF1 can be non-homologous in different mammalian species [1] , [60] , we used only the conserved region of ORF1 ( amino acids 123–321 , bp 1273–1869 of L1rp , GenBank accession number AF148856 ) as well as the region corresponding to full-length ORF2 of L1rp ( bp 1987–5814 ) for this comparison . The orthologous region of the reconstructed megabat ORF1 retains all the conserved amino acid sites , while the reconstructed ORF2 has two private changes ( L418V and V671T , bp 3238–3240 and 3997–3999 , respectively ) . These differences are consistent between our reconstruction and L1-2_PVa in RepBase and were verified in the original alignment to assure that they are not ambiguous in our reconstruction . We investigated the adenosine content of the reconstructed terminal members of megabat lineages 1 and 2 and 31 additional L1 consensus sequences from the mammalian species listed in Table S2 . L1 A-content of the two ORFs and the intergenic region ( IGR ) ranged from 39% to 44 . 5% , with a mean of 41 . 9% . Megabat L1 A-content was high among the species examined: lineage 1 ranked fifth at 43 . 7% and lineage 2 ranked second at 44 . 3% . To our surprise , the length of the megabat L1 IGR set it apart from the well-characterized L1s of rodents and primates . The IGR lengths of the surveyed L1 sequences from 31 species are listed in Table S2 and range from 18 to 580 bp . At 445 bp , the IGR of the reconstructed L1 is dramatically longer than either the median ( 63 bp ) or mean ( 172 bp ) among the species examined . Long IGRs were found among marsupials , Laurasiatheria ( which includes bats ) and Afrotheria species , but not among Euarchontoglires . Long IGRs are found in megabat families 1A ( 445 bp ) and 1B ( 481 bp ) , but the IGR length of families 2A ( 38 bp ) and 2B ( 26 bp ) is comparable to that of the majority of mammalian species . The IGR lengths in the remaining megabat L1 families are unknown . When multiple sequences were available in RepBase , we used the consensus of the most recently active L1 from each species for comparison; therefore , long IGRs could have existed in older or less active clades , or in sequences for which only partial reconstructions are feasible . To ask whether the reconstructed megabat L1 is capable of supporting retrotransposition , we synthesized it and assessed its activity in a retrotransposition rate assay derived from the work of Moran et al . [61] . This assay is routinely used to measure retrotransposition rates of L1s in a tissue culture system [47] , [62]–[64] . Reconstruction of fossil sequences can be challenging; even one error in reconstruction could block retrotransposition . Therefore , we synthesized the reconstructed gene in three segments and created all possible chimeric combinations using human L1rp [65]–[67] as a scaffold ( Figure 3 ) . Human L1rp is one of the most active natural human L1s characterized to date , and thus provides a robust background against which to test the effect of each L1 segment on retrotransposition rate . An independent L1rp construct , pWA192 [67] , was used as a positive control . An ORF1 mutant of L1rp [68] cloned in the same genetic context as the chimeric L1s was used as a negative control . The chimeric L1s are named by the source of their ORFs and IGR – H for human L1rp or B for the reconstructed megabat L1 . For example , HHH represents the two ORFs and IGR of L1rp ( GenBank accession number AF148856 ) , BBB represents the reconstructed megabat L1 ( GenBank accession number KF796623 ) and HBH represents the chimeric L1 that includes human ORF1 , megabat IGR and human ORF2 . Both reconstructed megabat ORFs support retrotransposition , but at lower rates than the highly active human L1rp ( Figure 4 ) . Comparisons between the human L1 ( HHH ) and the constructs containing either one or both of the megabat ORFs ( HHB , BHH and BHB ) show that replacing the human ORFs with a corresponding megabat version reduces the retrotransposition rate ∼26-fold . We note that the heterologous nature of the chimeric construct could be responsible for part of the retrotransposition rate reduction as shown by Wagstaff et al . [63] with the human-mouse chimeras . We verified retrotransposition in two positive colonies from each construct by ascertaining splicing of the G418 resistance intron by PCR using primers flanking the neo cassette ( Figure S2 ) . An alternative start codon for ORF2 , located in the IGR , would make ORF2 36 bp longer . We tested the retrotransposition rate of chimeric L1s based on this alternative ORF2 and no change in retrotransposition rate pattern was observed ( data not shown ) . The megabat IGR is inhibitory to retrotransposition . Replacing the native human L1 IGR with that of the reconstructed megabat ( HHH→HBH ) reduces the retrotransposition rate ∼26-fold , while introducing the human L1 IGR into the reconstructed megabat L1 ( BBB→BHB ) increases the retrotransposition rate ∼40-fold ( Figure 5A ) . In a mixed ORF context ( Figure 4B ) , both HHB→HBB and BHH→BBH result in ∼30-fold lower retrotransposition rates . Interestingly , the effect of the megabat IGR on the human construct ( HHH→HBH ) is similar to that seen when replacing either or both ORFs in the human construct with megabat ORFs ( HHH→HHB , BHH or BHB ) . The retrotransposition rates of the chimeric L1s are drastically lowered with the combination of the reconstructed megabat IGR and any of the reconstructed megabat ORFs ( BBH , HBB and BBB ) . Therefore , we conclude that compared to the HHH construct , the dampening effect of exchanging the ORFs is non-additive ( BHB vs . HHB and BHH ) , while exchanging either ORF and the IGR at the same time is approximately additive ( HHB vs . HBB , BHH vs . BBH and BHB vs . BBB ) . The hypothesis that retrotransposition rate is dependent on the amount of megabat L1 sequence in the construct is contradicted by the retrotransposition rate of BHB , which is largely made of megabat sequence but has a retrotransposition rate similar to those of constructs with only one bat segment ( HHB , BHH and HBH ) . To further investigate the inhibitory effect of the reconstructed megabat IGR on retrotransposition and its interaction with the L1 ORFs , we manipulated the megabat IGR and tested variants in the chimeric L1 context . Manipulation of the IGR included truncated versions of the full-length IGR , a shuffled version with the same nucleotide composition ( GenBank accession number KF796624 ) and an IGR with the sense-oriented AUG codons in all three reading frames mutated to AGU . We tested these variant IGRs in all four ORF contexts ( HXH , HXB , BXH and BXB , where X indicates the IGR variant ) . We found that while the absolute level of transposition was affected by whether human or megabats ORFs were framing the IGR , the relative decrease in retrotransposition was comparable in all ORF contexts . Therefore , the effect of the manipulated IGR on retrotransposition is shown only in the human L1rp context , HXH , in Figure 5C; the retrotransposition rates of the manipulated IGRs in all other ORF contexts are shown in Figure S3 . To determine whether the inhibitory property of the megabat IGR is due solely to its length , we truncated one-third or two-thirds of the IGR from either the 5′ end , the 3′ end or both ( Figure 5B ) . All the truncated IGRs increase the retrotransposition rate 0 . 3- to 0 . 5-fold compared to the full-length version ( Figure 5C; HBH compared to HaH , HbH , HcH and HbcH ) except the truncation of the 3′ one-third of the IGR ( Figure 5C; HBH compared to HabH ) , which decreases the retrotransposition rate ∼6 . 9-fold . Thus , while the length of the IGR accounts for part of its retrotransposition inhibition property , there are also effects from other factors . Although the megabat L1 IGR is inhibitory to retrotransposition compared to its human counterpart , we would expect to see that at this length , the reconstructed IGR still supports retrotransposition better than a randomized version with the same nucleotide composition . The randomized IGR with the same nucleotide composition reduces the retrotransposition rate ∼8 . 8-fold ( Figure 5C; HBH compared to HrH ) , suggesting that there is co-adaptation of the resident IGR with the L1 ORFs . Since it has been proposed that the translation of ORF2 is dependent on the existence of a close upstream ORF termination [4] , we expected to see lowered retrotransposition rates with all the small ORFs within the IGR eliminated , as this makes the stop codon of ORF1 the closest stop upstream of ORF2 and reduces the probability that ORF2 translation will reinitiate before the ribosome is released from the L1 transcript . Mutating the AUG codons in all three possible frames of the IGR into AGUs decreases the retrotransposition rate ∼3 . 3-fold compared to the intact bat IGR ( Figure 5C; HBH compared to H-H ) .
The acknowledged pattern of L1 evolution is that the active elements within a genome are closely related , giving rise to a single active lineage which dominates the total retrotransposition in the genome for a period of time [38] . Eventually the active elements accumulate debilitating mutations and become less active , but occasionally a new active element derived from an old one will emerge in the L1 population . This element can behave like a ‘stealth driver’ [69] and remain at low activity in the genome for a long period of time . When evolution drives a new element to high activity , the elements derived from it can eventually dominate the genome and give rise to a new family . Repetition of this lifecycle of L1 families results in the periodic fluctuation of L1 activity . Prior to L1 extinction , megabat L1s experienced periodic fluctuations in the number of elements fixed in the genome . This pattern is also observed in other mammalian clades , and in most cases each peak in copy number is dominated by a single L1 lineage . However , there are exceptions . For example , the human genome has been dominated by a single L1 lineage , but there was a period in primate evolution beginning about 46 MYA when two lineages were simultaneously active [35] . Similarly , two closely related lineages are currently active in the rodent genus Peromyscus [39] . Megabats stand out not only for the extinction of their L1s , but because their genomes have been continuously dominated by multiple active lineages with activity peaks of about the same age . Each peak includes two or three divergent families ( Figure 2 ) , a pattern that preceded the mammalian radiation and persisted throughout the history of L1 activity in megabats ( Figure 1 ) . Where multiple lineages are maintained , it is possible that they are specialized on different tissue types ( e . g . , on the male germ line vs . female germ line , or on the germ line vs . the embryo prior to differentiation of the primordial germ cells ) . Either of these scenarios could be successful in the evolutionary sense as mechanisms to avoid competition while still resulting in insertions that can be inherited by the next host generation . It is also possible that the L1 regulation mechanisms of the host are specific towards a certain lineage . Under that scenario , one lineage could dominate while the other is relatively quiescent , and eventually the second lineage could escape control and the first lineage be silenced . In other words , there would be no reason to expect that lineages would have the same peaks of increased retrotransposition . The fact that distinct lineages experienced fairly synchronized periods of activity and quiescence could suggest global rather than lineage-specific regulation of L1 retrotransposition . Peaks of L1 copy number are generally assumed to indicate transpositional bursts attributable to L1 activity , but other factors might account for peaks of L1 fixation in the genome . For example , host population bottlenecks could account for an increase in the rate of L1 fixation in the genome if there is selection against L1 [70] , and such bottlenecks would be expected to affect multiple lineages in a similar manner , accounting for simultaneous peaks of fixation . Another possibility is that these peaks are related to the propensity of L1s to insert into double-stranded breaks [47] , [51] , [71] , [72] . If a genome undergoes a period of extensive DNA damage due to an environmental or biotic assault , insertion into the resulting double-stranded breaks might lead to simultaneous peaks of retrotransposition of whatever L1 families are active at that time . To further characterize L1s in megabats at the time of their extinction , we reconstructed the full-length common ancestor of the most active family using a consensus-based method . Because of the unusual mode of L1 evolution [33] , [36]–[38] , consensus-based reconstruction is the preferred method of ancestral state reconstruction [56] , [73] . Reconstruction is particularly challenging for an extinct L1 family because of variation between old L1 insertions that have accumulated private mutations after elements inserted into the genome; this variation eventually dwarfs changes that occur as one family gives rise to the next , and thus to the phylogenetic signal relevant to evolution within active lineages . Since progeny of the most active elements within a family are over-represented in the genome , the resulting reconstructed sequence can best be thought of as representing the most active L1 master sequence at the time of L1 extinction . The reconstructed L1 sequence of megabat family 1A bears some of the features of a canonical L1 consensus from representative species , but also has some special characteristics to take into consideration . Although we identified and confirmed two amino acid changes in the reconstructed megabat ORF2 at sites that are conserved in all other species , such private changes at otherwise conserved sites were also frequently observed in the L1s used for comparison . The number of private changes in the L1s from a set of species varies from zero to seven with a median of two ( Table S2 and Text S1and S2 ) , which is in line with the number of private changes in the reconstructed megabat L1 . These same two changes were observed in the RepBase reconstruction , providing further confidence that they are not artifacts . It should be noted that mutations in this set of mammalian L1s are not totally saturated , so conserved sites are not necessarily functionally constrained , but functionally constrained sites should be among the conserved sites . Some sites likely appear to be conserved because of the limited number of ORFs available for comparison . An unusual aspect of L1 sequences is their high adenosine content on the coding strand and its possible dampening effect on transcription . This A-bias is prominent in the reconstructed megabat L1 , which ranks the fifth among the 31 species surveyed . For comparison , the adenosine content of the megabat genome trace file ( 30% ) is also slightly above the average level ( 29 . 5% ) of the species surveyed ( Table S2 ) . The A-richness of L1 can cause elongation [62] and post-transcriptional splicing defects [74] . It may also give rise to a codon usage pattern in L1s that is different from the codon usage of host genes . This implies that the high A-content of the reconstructed L1 is a possible contributor to its own retrotransposition rate and likely to have a dampening effect . It has been shown that A-bias correction with codon optimization increases the retrotransposition rate of a native , ‘hot’ mouse L1 by ∼200-fold [62] . Although the same optimization only increases retrotransposition rate of human L1rp ∼3-fold , the transcription of the codon-optimized L1rp is increased >40-fold [67] . The most unexpected feature of the reconstructed megabat L1 is its long IGR . Alisch et al . [4] and Li et al . [5] have shown independently that the IGR is indispensable for the translation of L1 ORF2 . The work of Alisch et al . [4] also demonstrated that the introduction of a long , structured IGR inhibits the retrotransposition of human L1s . This suggests that the long IGRs in megabat L1 lineage 1 may be inhibitory for retrotransposition . We cannot determine from examination of the megabat genome or from the work of Smit et al . [58] whether short or long spacers were ancestral among L1s of the Chiroptera ( bats ) . However , L1s with long IGRs can be found in some marsupials , Laurasiatheria and Afrotheria species . To determine whether the reconstructed megabat lineage 1 element was active , we made chimeric sequences using human L1rp , a highly active de novo insertion , as a backbone [65] , [66] . Ideally , these studies would have been carried out in both human and megabat cell lines . However , not all cell lines – and not all clones of permissive cell lines – support L1 retrotransposition . Megabat cell lines are not readily available , and we are unaware of an immortalized cell line from any bat that supports L1 activity . Fortunately , HeLa cells are competent hosts of heterologous and chimeric L1 retrotransposition . Mouse L1s readily retrotranspose in HeLa cells [75] , [76] as do chimeras between human and mouse L1s [63] . However , our studies differ from those of Wagstaff et al . [63] in that we did not codon optimize our L1 constructs . Although exchanging the L1rp ORFs with either or both of the corresponding megabat counterparts lowers the retrotransposition rate considerably , the activity of chimeric L1s is comparable to the majority of full-length human L1s . The retrotransposition rate of chimeric constructs containing megabat ORFs is much lower than the retrotransposition rate of the most active ‘hot’ L1s , but more active than 82% of full-length L1s in the human reference genome [41] . The retrotransposition rate of BBB is even lower , but still surpasses that of 56% of full-length L1s in the human reference genome . There are some caveats relevant to this comparison . First , the retrotransposition assays of Brouha et al . [41] were conducted in a different genetic background from the one in this study , but both studies use relative numbers normalized by the retrotransposition rate of L1rp , and thus are comparable . Secondly , although the reconstructed megabat L1 ( BBB ) supported retrotransposition at about the rate of the average active human L1 , it would not be expected to generate half the number of insertion events as a ‘hot’ human L1 because the contribution of individual active L1s to the total retrotransposition activity is unevenly distributed – just six ‘hot’ elements of the 80–100 full-length human L1s are responsible for more than 80% of the total retrotransposition activity [41] . Since the average human L1 barely contributes to the total L1 retrotransposition in the genome , we conclude that the intact reconstructed megabat L1 is able to retrotranspose , but by this measure transposes at a very low rate . The reconstruction did not include the promoter , as L1 retrotransposition driven by a native promoter is difficult to detect in tissue culture assays [64] . Therefore , interactions with heterologous regulatory sequences are not a factor in this assay . No single component of the reconstructed L1s was responsible for the inhibition of retrotransposition compared to L1rp; replacement of each component had a similar effect . This makes it unlikely that either a rate-limiting megabat L1 protein or an interaction with a specific host factor is responsible for dampening activity . We also note that these assays were conducted in a human cell line ( HeLa ) , which is heterologous to the reconstructed L1 , so these estimates must be interpreted with caution . Demonstrating activity of a reconstructed element in a tissue culture assay is the ultimate test of the quality of the reconstruction . To our knowledge , this is the first L1 element from a species that does not carry currently active L1s to be resurrected and tested for activity . However , ancestral L1s have been extensively reconstructed [58] and some of these reconstructions and their codon-optimized variants have been tested for activity in tissue culture assays . For example , Wagstaff et al . [73] showed that reconstructed ancestral L1 from primates are capable of retrotransposition . Another good example of a reconstructed ancient transposable element is Sleeping Beauty [77] , [78] , an element from fish which is active in human cells and has proven to be a powerful tool for genetic engineering . These reconstructed elements are ancient snapshots from lineages that have been co-evolving with a suite of host factors . It is important to remember that while we can reconstruct the sequence of the ancestral element , we cannot replicate the exact genetic context under which these reconstructed elements were active . RepBase have been actively reconstructing and hosting reconstructed ancestral transposable elements since its establishment [56] , [57] . However , detailed studies of the evolutionary history of a particular transposable element family usually focus on model organisms . The evolutionary history of human [35] and mouse [34] L1 lineages have been well-documented , but data are sparse for most mammalian clades . The work reported here complements that of Khan et al . [35] and Sookdeo et al . [34] , demonstrating the diversity of mammalian L1 evolution patterns and allowing us to understand mammalian L1 evolution at a broader level . The most striking feature of the reconstructed megabat L1 is the long IGR , which is is co-adapted with the ORFs to support retrotransposition . This is most evident in the comparison of the randomized IGR with the intact version ( Figure 5C ) , where retrotransposition with the intact IGR is 8 . 8-fold higher than the randomized version with the same base composition . Although the length of the IGR has a major effect on retrotransposition rate , other factors such as secondary structure and splicing sites of the L1 transcript can also dramatically change the retrotransposition rate . Li et al . [5] demonstrated that the IGR of a ‘hot’ mouse L1 , L1spa , contains an IRES that enhances the translation of a downstream ORF , and the work of Alisch et al . [4] suggests that the termination of another ORF directly upstream of the ORF2 start is the key for its translation . Our data demonstrate that the reconstructed L1 containing an AUG-codon-free IGR has a lower retrotransposition rate than that of the intact version . This is in line with the evidence found by Alisch et al . [4] as well as the original work by Horvath et al . [79] that proposes a reinitiation mechanism for the translation of dicistronic structures . Perhaps the most difficult aspect to reconcile about the long IGR in lineage 1 is its evolutionary persistence . An active element that deleted this long IGR would be expected to dramatically increase its retrotransposition rate and , thus , to dominate future retrotransposition . That is to say , there should have been strong selection favoring the deletion of the IGR . One might expect such a deletion to be ‘easy’ from an evolutionary perspective since it need not maintain a reading frame , and yet this did not happen . The tempo of L1 retrotransposition in megabats directly preceding L1 extinction is also noteworthy . A significant burst of retrotransposition occurred just prior to L1 extinction in megabats , contributing 25% of the detectable L1s to the genome . Family 1A accounts for the bulk of this activity – 18% of the total detectable elements in the genome – despite the demonstrated inhibitory effect of the long intergenic spacer on this family . The IGR has a long evolutionary history in this L1 lineage and likely preceded the evolution of megabats . Thus , despite its inhibitory effect on retrotransposition , it is unlikely that it contributed to L1 extinction . There are some characteristics of bat genomes that make them unique among the mammals . Bats , and especially megabats , have much smaller genomes than other mammals [80] . Data from 43 species of megabats , 62 species of microbats and ∼10 , 000 other mammalian species suggest that at 2 . 15 Gbp the megabat average genome size is significantly more constrained than the average of all mammals ( 3 . 42 Gbp ) and is considerably smaller than even the microbats ( 2 . 52 Gbp ) . It has been proposed that small genome size is related to the ability to fly given the high metabolic rate and small cell size requirements of flight [81]–[83] . For example , it has been shown that bird genomes are smaller and less variable in size than genomes of mammals and amphibians [80] and that their genome size is inversely correlated with their wing loading , an index of flight ability [84] . Since transposable elements are the major contributor to mammalian genome size [85] , pressure to constrain genome size will likely be reflected by stronger regulation of transposable elements . This regulation could theoretically result in both suppression of transposition and more efficient removal of inserted elements from the genome . Loss of L1 activity would be particularly effective in slowing expansion of the genome since L1s and the SINEs ( Short INterspersed Elements ) , that co-op the L1 replication machinery , together make up approximately a quarter of a typical mammalian genome [40] , [42] . Compared to other mammals , genome size constraint in bats confers a stronger selective pressure on the host defense mechanisms that control L1 retrotransposition , which could serve as the intrinsic driver for the host to develop anti-transposable element strategies that may increase the likelihood of transposable element quiescence and extinction in this group .
Since the large majority of L1s are truncated at the 5′ end [86] , the copy number of 3′ ends better represents the history of retrotransposition events . Therefore , we used 575 bp in the 3′ end of L1 ORF2 ( as reconstructed below ) to get a comprehensive view of L1 retrotransposition . Using the megabat L1 lineage 1 [50] consensus as the query sequence , we ran CENSOR 4 . 2 [87] against the ∼2× genome trace files of P . vampyrus ( Baylor College of Medicine , ftp . ncbi . nlm . nih . gov/pub/TraceDB/pteropus_vampyrus/ ) to find detectable sequences with >60% identity and >90% coverage of the query . Using 2000 random sequences from the CENSOR run , subfamilies were identified based on shared sequence variants ( co-segregating mutations ) with COSEG 0 . 2 . 1 ( http://www . repeatmasker . org/COSEGDownload . html ) [55] following the default parameters . Nine subfamilies were generated and their consensuses used as query sequences for a second round of CENSOR against the P . vampyrus genome . All identified L1 sequences from the second CENSOR run were used for a second round of COSEG , which required the additional parameter of at least 250 sequences to form a subfamily . Consensuses of the 64 subfamilies thus generated were used as query sequences to run CENSOR for a third time . Each hit's percent identity to the corresponding query was used to assign it to a L1 subfamily , and the copy numbers in each subfamily were counted . Seven subfamilies containing less than 250 sequences were removed . Consensuses from each of the remaining 57 subfamilies were used as query sequences to run CENSOR for a fourth time and all detected L1s were assigned to their subfamilies by the percent identity of each hit to its query . The 57 subfamily consensuses were aligned with ancestral mammalian L1s from RepBase [56] , [57] , reconstructed by Smit et al . [58] and Wade et al . [59] , with the Lasergene software suite ( DNASTAR , Madison , WI ) , and a distance matrix was calculated . Based on the alignment , a maximum likelihood tree was constructed using PhyML [88] with the GTR+I+G model and 100 bootstrap replicates ( Figure S1 ) . L1s were then assigned to families based on a <3 . 5% within-family pairwise distance from their subfamily consensuses . Sequence specificity of L1 families was determined by BLAST [89] against the NCBI whole genome sequencing databases . The consensus sequences of subfamilies 1 , 5 , 7 , 3 , 40 , 36 , 34 , 0 and 29 were used as the BLAST queries representing families 1A , 1B , 1C , 2A , 2B , 2C , 2D , 3 and 4 , respectively . A subfamily and its corresponding family were considered bat-specific only if <5 of the top 100 BLAST hits were not from bats . Histograms of L1 age distribution were generated by the R [90] histogram function using a window size of 0 . 5% ( Figures 1 and 2 ) . Percent identities corresponding to retrotransposition peaks of individual families ( Figure 2 ) were determined by R using the kernel smoothing function with 0 . 2% bandwidth . A full-length consensus sequence of the most recently active L1 from megabat lineage 1 was reconstructed by a series of progressive steps . The seed for the reconstruction was a conserved 575 bp region in the 3′ half of ORF2 ( Figure 3A ) . This region was previously amplified by degenerate PCR and a consensus sequence was determined [91] . Walks were performed in the 5′ and 3′ directions away from the cloned region and continued in both directions until full-length L1s were reconstructed . To aid with the reconstruction , a software pipeline was developed consisting of Perl ( http://www . perl . org/ ) , Ruby ( https://www . ruby-lang . org/en/ ) and Bash ( http://www . gnu . org/software/bash/ ) scripts . The pipeline queried , filtered and extracted data from the genome of P . vampyrus . An individual step resulted in the addition of 100–500 bp of sequence to the consensus , depending on the quality of the alignment at the ends , which was then used in the next step of the walk and in the final L1 reconstruction . Candidate sequences were identified in the database using BLAST with default parameters and an e-value of 1×10−50 , parsed through the BioPerl SearchIO module ( http://www . bioperl . org ) and screened based on their similarity to the input sequence . Only hits with at least 92% identity were retained to assure that the reconstruction did not include older lineages , and then a Ruby script extracted those sequences with overhangs of at least 100 bp . Alignments for each end were created and hand-edited to yield consensuses of clean read which were aligned into a master alignment . A 300–500 bp region from each end was selected to act as the seeds for the next step in the walk . The process was repeated until the entire element was reconstructed . Upon completion of the full-length L1 , a 500 bp seed was chosen arbitrarily from the final consensus and the pipeline was run again to verify the reconstruction . Methylated CpG sites evolve rapidly and must be corrected in the final consensus . CpG sites were identified by their high variation and the presence of dinucleotide sequence CG , CA , TG or TA; these were examined , manually edited and designated as CG in the final consensus . This pipeline also reconstructed the most recently active L1 lineage of four additional species listed in Table S1 , but required higher percent identities for the walks to reduce the noise introduced by older lineages . To compare the reconstruction of the extinct L1 to other L1s , sequences from a range of mammalian species were either reconstructed as described above , or selected from the RepBase report of February 2013 [56] . L1 consensuses of all species available in RepBase were aligned except those of dolphin and American opossum which had problematic regions of non-homology . When multiple L1 consensus sequences for the same species were present in RepBase , the one with highest average percent identity to its genomic sequence was chosen to represent the most recent master L1 in the genome . Some of the RepBase L1 sequences were out of frame at regions containing adenosine runs or contained in-frame stop codons , both resulting in significantly shorter ORFs . The following corrections brought these sequences into the correct reading frame: L1-1_Cpo , ignored an in-frame stop codon at bp 3050–3052 and used the original sequence for the alignment; L1-1_DV , added a N after bp 6015; and , L1A_Mim , deleted an A at bp 1590–1591 and bp 5336–5337 . The backbone plasmid for chimera constructions used in the retrotransposition assays was based on pL1PA1tag , a gift from Dr . Astrid Roy-Engel . pL1PA1tag contains a codon-optimized consensus of the PA1 family of human L1 in a pBSSK− ( Agilent Technologies , Inc . , Santa Clara , CA ) backbone . A puromycin resistance gene and its affiliated promoter pPGKpuro ( Addgene , Cambridge , MA ) were cloned into pL1PA1tag , creating plasmid pLY1004 . The L1 insert of pLY1004 was removed by NheI and EcoRI digestion , creating the final plasmid backbone ( Figures 3B and 3C ) . The reconstructed L1 and manipulated IGR sequences were commercially synthesized by GenScript USA , Inc . ( Piscataway , NJ ) . Reconstructed L1s were synthesized in two blocks consisting of ORF1+IGR and ORF2 . The manipulated IGRs were synthesized separately or in combinations containing distinct cloning sites . The synthesized sequences were cloned into pUC57 with flanking ends compatible to the linearized pLY1004 backbone and with BsaI or BsmBI sites to generate compatible overhangs after digestion . ORF1 and IGR were subcloned into separate pUC57 plasmids . Figure 3B illustrates the principle underlying the construction of the chimeric L1s . L1 ORFs and IGRs were amplified from these plasmids by PCR with Phusion high-fidelity polymerase ( ThermoFisher Scientific , Waltham , MA ) using primers designed to generate compatible overhangs when the PCR products are digested with BsaI , BtgZI or EcoRI . Human L1rp segments were cloned from pWA192 [67] , a gift from Dr . Wenfeng An , using the same principle . The L1 ORFs , IGRs and the linearized backbone plasmid pLY1004 were joined together by a multi-way ligation using T4 DNA ligase . All restriction enzymes and DNA modifying enzymes were from New England BioLabs , Inc . ( Ipswich , MA ) unless otherwise specified . All constructs were confirmed by sequencing the L1 insert . Retrotransposition rates were tested in an assay derived from Moran et al . [61] , in which the number of cell colonies surviving G418 antibiotic selection represents the retrotransposition rate ( Figure 3C ) . Briefly , the transcription and retrotransposition of L1 trigger the splicing of the transcript and excision of the intron of the inverse-oriented neo cassette , granting the cell resistance to the antibiotic G418 . The HeLa cell line ( ATCC CCL-2 ) was a gift from Dr . Wenfeng An and maintained in Dulbecco's Modified Eagle Medium with 4500 mg/L glucose and 110 mg/L sodium pyruvate ( ThermoFisher Scientific ) supplemented by 10% fetal bovine serum ( Atlanta Biologicals , Lawrenceville , GA ) , 2 mM l-alanyl-l-glutamine dipeptide and 100 units/mL Penicillin-Streptomycin ( ThermoFisher Scientific ) . The assay was conducted as described by An et al . [67] . The culture medium for antibiotic selection was similar to the cell maintenance medium except 2 . 5 ug/mL puromycin ( CALBIOCHEM , Billerica , MA ) or 50 mg/mL G418 ( CALBIOCHEM ) was added . Plasmids for transfection were prepared with the Promega ( Fitchburg , WI ) PureYield Plasmid Midiprep System and the cells were transfected with FuGENE HD transfection reagent ( Promega ) following the manufacturer's protocol . Retrotransposition assays of the chimeric L1s were repeated at least 12 times in three different batches and manipulated IGR assays were repeated at least four times . To confirm retrotransposition , two retrotransposition-positive colonies of each chimeric L1 construct were isolated with cloning rings , dissociated with trypsin ( ThermoFisher Scientific ) , seeded on T75 flasks and allowed to grow into confluence . Cells were harvested and their genomic DNA was extracted with the QIAamp DNA mini kit ( QIAGEN , Germantown , MD ) . Genotyping PCRs were conducted with primers bracketing the intron of the G418 reporter gene as described by An et al . [92] . Briefly , genotyping PCR primers were designed to the neo cassette so that cells hosting retrotransposition events , and the corresponding spliced cassettes , yield 653 bp PCR products . pLY1101 , a self-ligated version of the linearized pLY1004 without a L1 insertion , was constructed as a positive control; genotyping PCR of pLY1101 yields a 1556 bp construct corresponding to the unspliced neo cassette .
|
Most of a typical mammalian genome is occupied by transposable elements , which have played an important role in shaping these genomes , and L1s account for approximately half of this transposable element load . Mammals have evolved several mechanisms to control L1 retrotransposition , and yet L1s remain active in almost all mammalian lineages . However , L1s were found to have gone extinct in the megabat family ∼24 million years ago . We were able to trace megabat L1s to the ancestral L1 families shared by all mammals as well as identify bat-specific L1 families . Unlike most well-characterized mammals which have a single active L1 lineage , multiple L1 lineages have persisted in megabats throughout their evolutionary history . When the L1 extinction occurred in megabats , two active lineages lost their ability to retrotranspose almost simultaneously after a burst of activity . We synthesized the L1 from the most active family at the time of extinction and found a long intergenic spacer between its two protein coding genes . Tissue culture assays of the reconstructed megabat L1 revealed that both genes supported retrotransposition , but that the spacer is inhibitory . Despite the inhibition , this family accounted for 18% of the L1s detected in the megabat genome .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome",
"complexity",
"genomics",
"genome",
"evolution",
"genetics",
"synthetic",
"biology",
"biology",
"and",
"life",
"sciences",
"comparative",
"genomics",
"molecular",
"genetics",
"computational",
"biology",
"evolutionary",
"biology"
] |
2014
|
Reviving the Dead: History and Reactivation of an Extinct L1
|
It frequently has been postulated that intersexual coevolution between the male ejaculate and the female reproductive tract is a driving force in the rapid evolution of reproductive proteins . The dearth of research on female tracts , however , presents a major obstacle to empirical tests of this hypothesis . Here , we employ a comparative EST approach to identify 241 candidate female reproductive proteins in Drosophila arizonae , a repleta group species in which physiological ejaculate–female coevolution has been documented . Thirty-one of these proteins exhibit elevated amino acid substitution rates , making them candidates for molecular coevolution with the male ejaculate . Strikingly , we also discovered 12 unique digestive proteases whose expression is specific to the D . arizonae lower female reproductive tract . These enzymes belong to classes most commonly found in the gastrointestinal tracts of a diverse array of organisms . We show that these proteases are associated with recent , lineage-specific gene duplications in the Drosophila repleta species group , and exhibit strong signatures of positive selection . Observation of adaptive evolution in several female reproductive tract proteins indicates they are active players in the evolution of reproductive tract interactions . Additionally , pervasive gene duplication , adaptive evolution , and rapid acquisition of a novel digestive function by the female reproductive tract points to a novel coevolutionary mechanism of ejaculate–female interaction .
Extensive research across a broad range of taxa has revealed that the proteins involved in sexual reproduction often evolve rapidly due to positive selection ( reviewed in [1–3] ) . Although the selective forces that underlie this pattern remain unclear , it frequently has been postulated that adaptive evolution of reproductive proteins may result from intersexual coevolution [1–3] . Indeed , this has been demonstrated in the fertilization proteins of the free-spawning marine gastropod abalone , in which the male protein lysin and its female receptor , vitelline envelope receptor for lysin ( VERL ) , both exhibit signatures of adaptive evolution [4–7] . In internally fertilizing organisms , however , such as mammals or insects , the biochemical interactions between male and female reproductive proteins may be vastly more complex . Reproductive outcomes depend not only on interactions between male and female gamete proteins , but additionally on interactions between male seminal proteins and proteins in the lumen of a female's reproductive tract [8–11] . Fruit flies of the genus Drosophila provide an important model system for exploring the function and evolution of reproductive tract interactions ( reviewed in 9–12] ) . In Drosophila melanogaster , the male ejaculate comprises just under 100 proteins , several of which are known to stimulate important processes in mated females such as ovulation , oogenesis , and sperm storage ( reviewed in [9–11] ) . Several male proteins either undergo proteolytic cleavage in mated females [13–15] , or localize to specific portions of the female reproductive tract [16–18] , indicating that ejaculate–female interactions are mediated biochemically by females . Between species , rapid changes in ejaculate composition frequently have resulted in lineage-specific seminal proteins [19–21] , many of which may be novel coding sequences [22] . Additionally , molecular evolutionary studies indicate that a significant portion of this ejaculate is subject to positive selection in the melanogaster [23–25] , obscura [26] , and repleta species groups [27] . By comparison , the female side of reproductive tract interactions has received little attention . Female reproductive tract proteins have been identified transcriptionally only in D . simulans [28] , and their functions remain entirely unknown . Furthermore , although several female reproductive tract proteins [28–30] and egg membrane proteins [31] show evidence of positive selection , these analyses largely have been confined to the melanogaster species group . It is unclear , therefore , how diversity in female reproductive physiology and mating system across the genus [reviewed in 12 , 32] is reflected in their reproductive proteins . This overall paucity of research on females presents a major obstacle to understanding the evolution of ejaculate–female interactions and the role of intersexual dynamics in the divergence of reproductive proteins . Here we use a comparative expressed sequence tag ( EST ) approach to characterize candidate female reproductive tract proteins in D . arizonae . D . arizonae is a repleta group species that exhibits important differences from the melanogaster group in mating system and female physiology . D . arizonae females remate daily , while D . simulans females wait several days before remating [12] . Female promiscuity may affect the evolution of reproductive proteins by increasing the number of competing male ejaculates [33] . Females of D . arizonae additionally exhibit two remarkable post-mating physiological processes not seen in the melanogaster group . First , they incorporate peptide components of the male ejaculate into somatic tissues and oocytes [34] , an adaptation which may help defray the cost of egg production during periods of resource limitation [35] . Second , they exhibit an insemination reaction , an opaque white mass of unknown biochemical composition that forms in the female uterus after copulation [36] . By comparing post-mating outcomes in inter- and intrapopulation crosses , several studies have presented evidence for ejaculate–female coevolution in natural populations of D . arizonae and its sister species D . mojavensis ( most recent common ancestor , ∼1 . 5 million years ago [MYA] ) [37–41] . Intrapopulation crosses of both species produce larger eggs than interpopulation crosses [38] , a process known to be stimulated by several components of the male ejaculate in D . melanogaster ( reviewed in [9–11] ) . Additionally , the insemination reaction exhibits a larger size and duration in interpopulation crosses relative to intrapopulation crosses , suggesting this trait is subject to sexually antagonistic coevolution [39] . Finally , desiccation resistance is higher in mated than unmated females [40] , and the magnitude of this effect differs between inter- and intrapopulation crosses [41] . Such extensive evidence for physiological coevolution indicates this will be an exciting system to explore the molecular basis of reproductive tract interactions . Our study identifies 241 candidate female reproductive proteins in D . arizonae , of which 31 show elevated rates of amino acid substitution suggestive of adaptive evolution . Unexpectedly , we also discovered three lineage-specific gene families of digestive proteases whose expression is specific to the lower female reproductive tract . These proteins exhibit strong signatures of adaptive evolution , and selected sites cluster near functionally important amino acids . The implications of these findings for ejaculate–female interactions and intersexual coevolution are discussed .
We sequenced a total of 2 , 304 ESTs derived from the D . arizonae lower female reproductive tract ( parovaria , oviduct , spermathecae , seminal receptacle , and uterus ) representing 649 unique proteins ( for a complete list see Table S1 ) . Of particular interest are proteins found on cell surfaces or in the lumen of this tissue , which interact directly with the male ejaculate and likely play an integral role in reproductive tract interactions [28] . We therefore designate candidate female reproductive proteins as those that exhibit secreted signal sequences , or transmembrane domains . The gross functional composition of the 241 candidate female reproductive proteins identified in this study ( Figure 1 ) are similar to those of D . simulans [28] , and include transport , signal transduction , and proteolysis . To explore the evolutionary histories our candidate female reproductive proteins , we calculated the ratio of replacement to silent substitutions ( dN/dS ) between our D . arizonae ESTs and their orthologs in the D . mojavensis genome . Candidate female reproductive proteins exhibit significantly larger dN/dS values than intracellular proteins in our dataset ( median test , p > 0 . 0001 ) , suggesting that these proteins evolve more rapidly than their intracellular counterparts . This elevated rate of amino acid substitution is predicted if adaptive evolution of secreted and transmembrane proteins is a frequent consequence of molecular coevolution with components of the male ejaculate . Under strict neutrality , only dN/dS ≫ 1 can be considered robust evidence of adaptive evolution . While several of our candidate genes show dN/dS > 1 , none of these tests is statistically significant ( Table 1 ) . A literature survey has shown , however , that 95% genes that exhibit a pairwise dN/dS > 0 . 5 contain a class of sites with dN/dS ≫ 1 [28] . Of 227 pairwise comparisons , 31 ( 14% ) were identified with dN/dS > 0 . 5 , indicating they are likely experiencing positive selection ( Table 1 ) . This result is largely independent of gene duplication , as the estimated frequency of adaptive evolution it is still 13% when recent duplicates are excluded from the dataset . On a functional level , several protein classes that commonly occur in seminal and fertilization proteins , including lipases , lectins , glycoproteins and proteases , are found in our candidates for adaptive evolution ( Table 1 ) . Roughly half of these 31 candidates , however , have no known function , and several others belong to functional classes that are not commonly represented among reproductive proteins . Proteins with unusual or unknown functions make excellent candidates for discovering genes which have acquired novel functions in a biochemical network which likely evolves rapidly . Future studies of these 31 candidates will yield significant insight into the function and evolution of reproductive tract interactions in the repleta species group . Gene duplication plays an integral role in the evolution of D . arizonae female reproductive tract proteins . Specifically , 47% ( 16 ) of all secreted proteases in D . arizonae female reproductive tracts have at least one closely related paralog that also is expressed in these same tissues . Duplication events have been extremely recent; as multiple , tandemly-duplicated paralogs in the D . mojavensis genome correspond to only a single gene in D . virilis , the most closely related fully sequenced outgroup ( most recent common ancestor , ∼23 MYA; reviewed in [42] ) . We therefore estimate that the duplication rate of secreted proteases expressed in D . arizonae tracts is 0 . 0298 ( duplications per gene per million years , see Materials and Methods ) , which is 21-fold higher than the genome wide estimate for D . melanogaster ( 0 . 0014 , [43] ) . Although the selective forces involved are yet obscure , such recent and pervasive gene duplication has not been seen in any class of reproductive protein yet studied , including D . simulans female reproductive proteins [28] . Four ( of 16 ) duplicated proteases have resulted from two single gene duplication events . The remaining 12 duplicated proteases , however , are associated with small lineage-specific gene families . Each family contains four to six tandemly duplicated paralogs in the genome of D . mojavensis that are syntenic to a single ortholog in the genome of D . virilis ( Figure 2 ) . For brevity , we hereafter refer to these three families of tandem duplicates as protease gene family 1 , 2 , and 3 . Phylogenetic analysis of D . arizonae ESTs , and coding sequences from the genomes of D . mojavensis , D . virilis , and D . grimshawi ( http://rana . lbl . gov/drosophila ) , reveals the majority of these tandem duplicates in the D . mojavensis genome have a D . arizonae ortholog that is expressed in the lower female reproductive tract ( Figure 3 ) . This strongly suggests that the gene duplication events relate in some way to the reproductive function of these proteases . Indeed , reverse transcriptase PCR ( RT-PCR ) of all three gene families reveals that in adult D . arizonae these genes are exclusively expressed in the lower female reproductive tract ( Figure 4 ) . Gene copies present in the D . mojavensis genome that do not correspond to D . arizonae ESTs are likely not highly expressed . While the function of these duplicated proteins in D . arizonae female reproductive tracts is unknown , they are often similar or identical in their key amino acid residues to several families of digestive proteases found almost exclusively in gastrointestinal tracts ( Table 2 ) . Specifically , protease gene families 1 and 2 share appreciable homology with trypsin , chymotrypsin , and elastase , serine endopeptidases commonly found in digestive tracts of both insects and mammals [reviewed in 44] . While , serine endopeptidases can also function in immune signaling cascades across a broad array of organisms , such proteases generally have secondary protein–protein interaction domains that allow for localized regulation of physiological responses [45] . No such domains are seen in either protease gene family 1 or 2 , suggesting these proteases exhibit a primarily digestive function . Similar to the two families of serine endopeptidases , protease gene family 3 contains zinc metalloendoproteases very similar to astacin , a prominent digestive enzyme in the crayfish midgut [reviewed in 46] . The reproductive tract-specific expression of these proteases , coupled with recent , lineage-specific gene duplications , suggest that D . arizonae female reproductive tracts recently have acquired a novel digestive function . Digestive enzymes in female reproductive tracts likely have important implications for male reproductive success , and therefore , the evolution of the male ejaculate . There is compelling evidence that directional selection has played an important role in the evolution of reproductive tract-specific secreted digestive proteases in D . arizonae females . All three families of digestive proteases exhibit a class of sites whose ratio of nonsynonymous to synonymous substitutions ( dN/dS ) is significantly greater than the neutral expectation of 1 ( Table 2 ) . dN/dS values for these selected sites range from 2 to 11 . 96 , indicating certain amino acids in these proteins have experienced strong positive selection . Notably , the two single gene duplication events show no evidence of adaptive evolution ( Table 2 ) , indicating that directional selection has been exclusive to the lineage-specific families of digestive proteases . In order to interpret selection in terms of both duplication and speciation events , we used the PAML free ratios model [47] to estimate dN/dS along every branch in each of the three phylogenies ( Figure 3 ) . Positive selection associated with three different speciation events suggests that ongoing changes in the biochemical environment of the female reproductive tract , including possible male contributions to this environment , have resulted in adaptive evolution in some of these proteins . A total of five gene duplication events are also immediately followed by a period of positive selection in one of the paralogous branches ( dN/dS > 1 ) , indicating neofunctionalization of a duplicate gene copy . The other seven duplication events however , are followed by elevated amino acid substitution rates ( dN/dS = 0 . 2–1 ) but no evidence of adaptive evolution . This suggests that relaxed constraint created by functional redundancy between paralogs has also played an important role in the evolution of these gene families . Evidence for adaptive amino acid evolution in duplicated genes implies that selection has acted to diversify the paralogs functionally . Indeed , in all three of the protease gene families , polar , nonpolar , and charged amino acids are seen to inhabit the same selected site in different paralogs . This indicates that directional selection has resulted in recurrent and radical amino acid substitutions , likely affecting the structure and function of the encoded proteins . By mapping selected sites onto predicted molecular structures , it is possible to make more specific inferences about how the biochemical function of these enzymes has been impacted by adaptive evolution . In the two families of serine endopeptidases ( protease gene families 1 and 2 ) , positive selection clusters near the catalytic triad: the three amino acids essential for proteolytic function ( reviewed in [44] ) ( Figure 5 ) . Furthermore , in protease gene family 1 , positive selection is found adjacent to , and in one case synonymous with , three amino acid sites known to effect substrate specificity ( reviewed in [48] ) . Collectively , these data indicate that directional selection has acted to diversify the catalytic activity of both families of serine endoproteases , and that protease gene family 1 has concomitantly undergone adaptive evolution for increased breadth in substrate specificity . Future functional studies of these enzymes , particularly in terms of how they interact with the male ejaculate , will yield significant insight into the selective pressures that underlie diversification of these extraordinary gene families . Our most striking result was the observation of three lineage-specific radiations of secreted digestive proteases in D . arizonae female reproductive tracts . Although the biological significance of these gene duplications is yet unclear , they may relate to two unusual physiologies exhibited by both D . arizonae and D . mojavensis females . First , the insemination reaction must be degraded by females prior to oviposition or remating [36] , a process that could require specialized digestive machinery . Second , female incorporation of ejaculate-derived protein , as observed in D . arizonae and D . mojavensis , could be facilitated by degrading seminal proteins and/or sperm into smaller fragments that are more easily absorbed . Regardless of their physiological function , lower female reproductive–tract specific expression of digestive enzymes points to a novel form of ejaculate–female interaction , in which females may actively degrade , rather than process or activate [13–15] , protein components of the male ejaculate . Digestion of seminal proteins or sperm would undoubtedly have important implications for male reproductive success , predicting an evolutionary response from males . Indeed , the association of these proteases with recent gene duplications and strong signatures of adaptive evolution suggests they are involved in an intersexual arms race . Exploring the male side of this interaction , therefore , is an important avenue of future research . The 31 candidates for adaptive evolution also have important implications for reproductive tract interactions and intersexual coevolution . Roughly half of these proteins have no known function or conserved domain , suggesting they are enriched for novel biochemical functions . Additionally , the candidates include several classes of proteins that have not been implicated previously in reproductive tract interactions . Particularly intriguing are three transmembrane proteins with the conserved transporter domain MFS_1 , for inorganic solutes ( Table 1 ) . Although the biochemical composition of the Drosophila ejaculate is largely unknown outside of its protein constituents , females of several species incorporate ejaculate-derived phosphorus into somatic tissues and oocytes [49] . It is unclear if these transporters underlie such a process in D . arizonae . Their presence and evolutionary history point , however , to nonpeptide biochemical interactions in female reproductive tracts which also may evolve rapidly . If divergence of reproductive proteins is driven by intersexual dynamics , particularly sexually antagonistic coevolution [50–52] , species with more promiscuous mating systems are predicted to exhibit comparatively more adaptive evolution in their reproductive proteins . D . arizonae is significantly more promiscuous than its previously examined congener D . simulans [28] , and , consistent with the prediction , we find evidence that this difference in mating system may be reflected in the evolution of their female reproductive proteins . Specifically , we observed that candidate female reproductive proteins in our dataset exhibit higher dN/dS values than intracellular proteins , while this effect was not seen in similar comparisons between D . simulans and D . melanogaster [28] . Additionally , the estimated frequency of adaptive evolution in D . arizonae female reproductive tract proteins ( 14% ) is significantly higher ( Fisher's Exact Test p = 0 . 003 ) than that of D . simulans ( 5% ) [28] . Although the experimental approach for these two studies was quite similar , differences in divergence times between D . arizonae and D . mojavensis ( ∼1 . 5 MYA , [37] ) , and D . simulans and D . melanogaster ( ∼3 MYA , [53] ) , could result in more stochastic influence on our measures of dN/dS . Firm conclusions about the effect of mating system on the evolution of female reproductive proteins therefore requires further empirical testing across a broader array of taxa . Although the function and evolution of male seminal proteins have been researched extensively in both insects and mammals , our understanding of the female reproductive tract proteins with which they interact remains sparse . Our data , as well as previous research in the melanogaster group [28–30] , indicate that rapid evolution is common among female reproductive tract proteins . We furthermore present compelling evidence that differences in female physiology and possibly mating system between Drosophila species are reflected in their reproductive tract proteins . Our research indicates that female reproductive proteins are active players in reproductive tract interactions , and that rapid evolution of seminal proteins must be considered in terms of their relationship with female counterparts .
D . arizonae used in this study were collected in December 2005 in Tucson , Arizona by E . S . K . A total of 873 lower reproductive tracts ( parovaria , oviduct , spermathecae , seminal receptacle , and uterus ) were dissected from mature adult females 9 d or older . In order to maximize transcriptional diversity obtained , dissected females were sampled from a diverse array of mating states . Of the females , 662 were from population bottles , while approximately 40 females were dissected from each of the following treatments: virgin , homospecifically mated 4–8 h postcopulation , homospecifically mated 24 h postcopulation , heterospecifically ( to D . mojavensis ) mated 4–8 h postcopulation , and heterospecifically mated 24 h postcopulation . The harvested tracts were pooled into four separate aliquots of TRIZOL reagent ( Invitrogen , http://www . invitrogen . com ) and total RNA was extracted according to manufacturer instructions . Quality of these samples was verified with an Agilent 2100 bioanalyzer ( http://www . home . agilent . com/ ) , at which point they were pooled . mRNA enrichment was achieved by binding poly-A tails on Oligotex ( Qiagen , http://www . qiagen . com/ ) spin columns . Quality of enriched mRNA was verified with an Agilent 2100 bioanalyzer , and the total yield ( 1 . 5 μg ) was used for library construction with the Cloneminer cDNA library construction kit ( Invitrogen ) . Approximately 300 , 000 colony-forming units were obtained with an estimated insert size of 1kb . Of these clones , 10 , 000 were picked with a QBOT ( Genetix , http://www . genetix . com/ ) operated by the Arizona Genomics Institute ( http://www . genome . arizona . edu/ ) . Of these clones , 1 , 920 were sequenced bidirectionally , and an additional 384 were sequenced exclusively from their 5′ ends . All sequencing was done on at the Arizona Genomics Institute on an ABI 3700 DNA analyzer ( https://products . appliedbiosystems . com/ ) with big-dye terminator chemistry . Base calling and assembly were implemented in Phred and Phrap [54] . All bases with a Phred quality score below 20 ( 99% accurate ) were excluded from further analysis . The estimated frequency of sequencing errors in included bases was 0 . 04% . BLASTN [55] ( e-value = 0 . 01 ) against the GLEANR coding sequence annotations ( from CAF1 assembly http://rana . lbl . gov/drosophila/ ) of the D . mojavensis genome was used to identify orthologs of D . arizonae ESTs . For ESTs with no good BLASTN hit to annotated coding sequence , BLASTN ( e-value = 0 . 01 ) was implemented against the complete CAF1 assembly of the D . mojavensis genome . ESTs with BLAST hits in the D . mojavensis genome that contained long open reading frames were used to annotate additional genes in D . mojavensis by eye . No examples of ESTs with long open reading frames but no good BLASTN hit in the D . mojavensis genome were identified . Translations of these coding sequences were used to identify secreted proteins and cell surface receptors using SignalP [56] , and transmembrane proteins using TMHMM [57] . Conserved protein family ( Pfam ) domains were identified with hmmpfam [58] . Gene Ontology ( GO ) terms [59] were obtained from FlyBase ( http://flybase . bio . indiana . edu/ ) for D . melanogaster homologs , or based on conserved Pfam domains if no D . melanogaster homolog was found . For explicit definitions of GO terms see http://www . geneontology . org/ . In total , the D . arizonae ESTs corresponded to 649 unique proteins in the D . mojavensis genome . The orthologous genes were aligned using CLUSTALW [60] and alignment accuracy was verified by eye . Maximum-likelihood estimates of nonsynonymous substitutions rate ( dN ) , synonymous substitution rate ( dS ) , and the ratio of nonsynonymous substitutions per nonsynonymous site to synonymous substitutions per synonymous site ( dN / dS ) , were obtained from PAML [47] . For duplicated genes , only reciprocally monophyletic homologs were compared in pairwise analyses . Sequence data for D . arizonae was obtained from the EST library , while sequences from D . mojavensis , D . virilis , and D . grimshawi were obtained from their unpublished , publicly available genomes ( http://rana . lbl . gov/drosophila/ ) . GENECONV was used to test for gene conversion between paralogs , using the method of Sawyer [61] . Phylogenetic reconstruction of multigene families was implemented in Mr . Bayes v3 . 0b4 . Nested maximum-likelihood models of codon evolution were implemented in the codeml program of PAML [47] and compared using likelihood ratio tests . Two tests of positive selection were performed . In the first test , the neutral model ( M1 ) is compared with the selection model , in which a class of sites is permitted to exhibit dN/dS ( ω ) > 1 ( M2 ) . In the second test , a beta distribution of site classes in which the most rapidly evolving is fixed to ω = 1 ( M8a ) is compared to a similar model in which the most rapidly evolving site class is permitted to exhibit ω > 1 ( M8 ) [62] . Multiple initial values of ω were used to ensure convergence on the likelihood optima . For the second test , critical values of the test statistic are determined from Wong et al [63] . Lineage-specific selection patterns of dN/dS were determined by implementing branch-specific models [64] . A total of 34 secreted proteases were identified in D . arizonae female reproductive tracts . Using BLASTN homology and maximum-likelihood phylogenetic reconstruction implemented in PAUP* , we determined these 34 proteins correspond to 37 orthologs in the genome of D . mojavensis , and 23 orthologs in the genome of D . virilis ( http://rana . lbl . gov/drosophila/ ) . Assuming no gene conversion or gene loss , the total copy number of these genes was 23 at the divergence of the D . mojavensis and D . virilis lineages . Duplication rate can therefore be estimated by the following exponential growth equation: Where CM is copy number of D . mojavensis ( 37 ) , CA is the ancestral copy number ( 23 ) , t is the divergence time between D . mojavensis and D . virilis ( t = 23 MYA [42] ) , and r is the estimated rate of duplication per gene per million years . D . arizonae RNA was extracted from 20 whole males , 70 reproductively mature females from population bottles lacking their lower reproductive tracts , and 70 lower reproductive tracts preserved in TRIZOL ( Invitrogen ) according to manufacturer instructions . Purified RNA was treated with DNAseI ( Gibco , http://www . invitrogen . com/ ) , and reverse transcribed with the iScript cDNA synthesis kit ( Bio-Rad , http://www . bio-rad . com/ ) . Resultant cDNA was diluted to 10 ng/μl , and used as a template for standard PCR using universal primers , with D . arizonae genomic DNA as a positive control . Primer sequences are as follows: Dmoj\GLEANR_8528-F , 5′-AAGAAGCGCACCAAGCACTTCATC-3′; Dmoj\GLEANR_8528-R 5′-TCTGTTGTCGATACCCTTGGGCTT-3′; protease gene family 1 -F1 5′-ATGTGGAATCTAAGCCCAGCCAA-3′; protease gene family 1 -F2 5′-RTAGATGGCAGTTGCTYCTYGTG-3′; protease gene family 1 -R1 5′-GATGYGATACCAATCACRGTGCT-3′; protease gene family 1 -R2 5′-ACGATRCCAATCACRGTGCYAGA-3′; protease gene family 2 -F1 5′-CTCAAACCGCARTAGYTRTCCT-3′; protease gene family 2 -F2 5′-CTTCAAGCCGCMGTWGCTGTCCT-3′; protease gene family 2 -R1 5′-CACCRCTGTGYTYCCTRATCCATTC-3′; protease gene family 2 -R2 5′-CACCGCWGTGCTCYYTGATCCATT-3′; protease gene family 3 -F1 5′-TGAAACCGATCCCAGACTTATAGC-3′; protease gene family 3 -F2 5′-ATGAAACCGATCCCGAGTTGATAG-3′; protease gene family 3 -R1 5′-ATCAGCCATGCTCAATTCTTGTCG-3′; and protease gene family 3 -R2 5′-ATCAGCCCAGCTTAATTCTAGTCG-3′ . Three dimensional structure was predicted by SWISS-MODEL [65] , and visualized by Deep View . Selected sites were determined from Bayes Emperical Bayes calculation [66] implemented under M8 in PAML [47] .
All sequences for this study are available from the National Institute for Biotechnology Information ( NCBI ) GenBank ( http://www . ncbi . nlm . nih . gov/Entrez/index . html ) accession numbers EV41299147751410–EV41383447752253
|
In a broad range of organisms , including humans , molecular interactions between the male ejaculate and the female reproductive tract play integral roles in sexual reproduction . Although these interactions are essential , the biochemical composition of the male ejaculate can change rapidly over short evolutionary time periods . It is often hypothesized that this rapid evolution reflects a coevolutionary relationship with the female reproductive tract . The paucity of research on females , however , presents a formidable challenge to empirical tests of this hypothesis . In this study , we sought to identify proteins in the female reproductive tracts of D . arizonae that may be interacting or coevolving with the male ejaculate . Unexpectedly , we discovered that D . arizonae females produce an array of “digestive” enzymes in their reproductive tracts . These classes of enzymes are normally found in the gut , where they degrade ingested food for nutritional uptake . In D . arizonae , these enzymes have resulted from recent gene duplications , and natural selection has caused rapid and radical changes in their amino acid sequences . We propose that this pattern of duplication and diversification reflects the “female side” of a coevolutionary relationship with the male ejaculate . Exploring the “male side” of this relationship is an important avenue for future research .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"evolutionary",
"biology",
"drosophila",
"physiology"
] |
2007
|
Gene Duplication and Adaptive Evolution of Digestive Proteases in Drosophila arizonae Female Reproductive Tracts
|
Enteroaggregative E . coli ( EAEC ) have been associated with mildly inflammatory diarrhea in outbreaks and in travelers and have been increasingly recognized as enteric pathogens in young children with and without overt diarrhea . We examined the risk factors for EAEC infections and their associations with environmental enteropathy biomarkers and growth outcomes over the first two years of life in eight low-resource settings of the MAL-ED study . EAEC infections were detected by PCR gene probes for aatA and aaiC virulence traits in 27 , 094 non-diarrheal surveillance stools and 7 , 692 diarrheal stools from 2 , 092 children in the MAL-ED birth cohort . We identified risk factors for EAEC and estimated the associations of EAEC with diarrhea , enteropathy biomarker concentrations , and both short-term ( one to three months ) and long-term ( to two years of age ) growth . Overall , 9 , 581 samples ( 27 . 5% ) were positive for EAEC , and almost all children had at least one detection ( 94 . 8% ) by two years of age . Exclusive breastfeeding , higher enrollment weight , and macrolide use within the preceding 15 days were protective . Although not associated with diarrhea , EAEC infections were weakly associated with biomarkers of intestinal inflammation and more strongly with reduced length at two years of age ( LAZ difference associated with high frequency of EAEC detections: -0 . 30 , 95% CI: -0 . 44 , -0 . 16 ) . Asymptomatic EAEC infections were common early in life and were associated with linear growth shortfalls . Associations with intestinal inflammation were small in magnitude , but suggest a pathway for the growth impact . Increasing the duration of exclusive breastfeeding may help prevent these potentially inflammatory infections and reduce the long-term impact of early exposure to EAEC .
Enteroaggregative Escherichia coli ( EAEC ) infections have been increasingly recognized as important enteropathogens since their initial discovery by patterns of adherence to HEp-2 cells in E . coli isolates from Chilean children with diarrhea [1] . EAEC have since been associated with foodborne outbreaks of diarrhea [2] , traveler’s diarrhea [3–5] , diarrhea in adults with HIV infection [6] , endemic diarrhea in cities in the US [7] , and variably in healthy adult human volunteers [8 , 9] . A meta-analysis of 41 studies found EAEC to be significantly associated with acute diarrheal illness among both children and adults in developing regions [10] . However , because EAEC are also a highly common infection among children without overt diarrhea in low-resource settings , they have not been found to be a major cause of diarrhea in some endemic settings [11 , 12] . Regardless , EAEC , independent of diarrheal symptoms , have been associated with other poor health outcomes in children , such as growth failure [13] and mild to moderate intestinal inflammation [5 , 13 , 14] . The genetic determinants and biological mechanism for the virulence of EAEC have been described by a complex array of interacting traits that reside on both the chromosome and plasmid in the organism [15 , 16] . As presently defined , EAEC are heterogeneous with respect to virulence gene content . The aggR trait on the plasmid is a common and well-characterized EAEC gene [17] that regulates many virulence traits , including chromosomal aaiC , which is in the gene cluster aaiA-Y that encodes the type VI secretion system , as well as plasmid-borne aatA , which encodes an ABC transporter . In addition , the flagellin of EAEC strain 042 has been shown to trigger inflammation via TLR5 signaling [18 , 19] . Murine models have helped determine the impact of these virulence genes by providing evidence that EAEC can cause inflammation , enteropathy , and growth shortfalls among mice with dietary protein deficiency [20 , 21] , and even diarrhea among mice with dietary zinc deficiency [22] . Increasing evidence suggests that enteric infections , especially common pathogens like EAEC , may play an important role in morbidity due to enteric disease , beyond symptomatic diarrhea [23] . While mortality from diarrheal diseases has been dramatically reduced to less than half a million deaths per year [24] , more than a quarter of the world’s children are moderately or severely stunted [25] . Because improved feeding does not eliminate growth shortfalls in low-resource areas where inadequate water and sanitation and heavy burdens of enteric infections are common [26 , 27] , enteric infections and sub-clinical environmental enteropathy likely also contribute to poor child growth outcomes [28 , 29] . We characterized the epidemiology and impact of EAEC infections among children in the first two years of life in eight low-resource settings of the Etiology , Risk Factors , and Interactions of Enteric Infections and Malnutrition and the Consequences for Child Health and Development Project ( MAL-ED ) study . With twice-weekly active surveillance from near birth to two years of age , the MAL-ED study provides a unique opportunity to assess the impact of both clinical and subclinical enteric infections on early-life growth and development . We examined risk factors for EAEC infections and their associations with diarrhea , environmental enteropathy biomarkers , and growth outcomes over the first two years of life .
The study was approved by the Institutional Review Board for Health Sciences Research , University of Virginia , USA as well as the respective governmental , local institutional , and collaborating institutional ethical review boards at each site: Ethical Review Committee , ICDDR , B ( BGD ) ; Committee for Ethics in Research , Universidade Federal do Ceara; National Ethical Research Committee , Health Ministry , Council of National Health ( BRF ) ; Institutional Review Board , Christian Medical College , Vellore; Health Ministry Screening Committee , Indian Council of Medical Research ( INV ) ; Institutional Review Board , Institute of Medicine , Tribhuvan University; Ethical Review Board , Nepal Health Research Council; Institutional Review Board , Walter Reed Army Institute of Research ( NEB ) ; Institutional Review Board , Johns Hopkins University; PRISMA Ethics Committee; Health Ministry , Loreto ( PEL ) ; Ethical Review Committee , Aga Khan University ( PKN ) ; Health , Safety and Research Ethics Committee , University of Venda; Department of Health and Social Development , Limpopo Provincial Government ( SAV ) ; Medical Research Coordinating Committee , National Institute for Medical Research; Chief Medical Officer , Ministry of Health and Social Welfare ( TZH ) . Informed written consent was obtained from the parent or guardian of each participating child on their behalf . We identified risk factors for EAEC detection in surveillance stools using log-binomial regression with general estimating equations ( GEE ) and robust variance to account for correlation between stools within children , adjusting for site and a restricted quadratic spline [38] for age . Variables were assessed individually in this model and were included in the multivariable model if statistically significant ( p<0 . 05 ) . We estimated the association between EAEC and diarrheal versus non-diarrheal stools using Poisson regression with the robust variance estimator to estimate risk ratios [39] since log-binomial models did not converge , adjusting for the age spline , site , the interaction between age and site , and antibiotic use within the preceding 15 days . We then estimated the association between EAEC detection and stool biomarker concentrations ( ALA , MPO , and NEO ) on the logarithmic scale in the same stool using multivariable linear regression with GEE and robust variance to account for correlation between stools within children . We also estimated the association of EAEC detection with serum and urine biomarkers ( AGP and LMZ , respectively ) measured in the same month as the stool collection . Because Campylobacter was the most common pathogen detected in stools and has been previously shown to be associated with intestinal inflammation in the MAL-ED cohort [40 , 41] , we assessed potential interactions between the effects of EAEC and Campylobacter on MPO by including an interaction term between presence of EAEC and Campylobacter . All estimates were adjusted for site , the age spline , sex , WAMI , percent exclusive breastfeeding in previous month , contemporary presence of Campylobacter in the stool sample , and a qualitative description of stool consistency ( for stool biomarkers only ) . Finally , we estimated the association between EAEC detection and short-term and long-term growth using multivariable linear regression . Short-term growth was defined by the change in WAZ and LAZ over both the one and three months following each monthly stool collection . We compared differences in short-term growth velocity between children who had surveillance stools with and without EAEC detection , using GEE and adjusting for site , age , sex , WAMI , percent exclusive breastfeeding in the exposure month , and detection of Campylobacter in the stool . We further assessed the interaction between MPO levels and EAEC positivity to explore the role of intestinal inflammation in the potential effect of EAEC on short-term growth impairment . In the adjusted short-term growth models examining WAZ and LAZ velocity over the one and three months following EAEC testing , we estimated the additive interaction effect of EAEC detection and high MPO concentration in the same stool using an interaction term . High MPO was defined as an MPO concentration in the highest quartile on the logarithmic scale among all non-diarrheal stools collected at that child’s site and 3-month age period . Values defining high MPO ( range: 2 , 515–33 , 190 ng/mL ) were higher than previous reports from non-tropical settings ( <2 , 000 ng/mL ) [42] . Effects on long-term growth were then estimated as the total difference in size at two years of age as a function of the percent surveillance stools positive for EAEC . The long-term model was adjusted for the WAZ and LAZ measurements at enrollment ( within 17 days of birth ) , site , sex , WAMI , the age at which exclusive breastfeeding first stopped , and the percent surveillance stools positive for Campylobacter in the first 2 years of life . Adjusting for the same covariates , we assessed the potential synergistic interaction between the effects of EAEC and Campylobacter on growth at 2 years given that both have been associated with gut inflammation , by including an interaction term between an indicator for a high frequency of detection ( at least 50% surveillance stools positive ) of EAEC and an indicator for a high frequency of detection of Campylobacter . We also repeated the model described above , but focused on EAEC detections in specific age periods ( 1–6 , 7–12 , and 15–24 months ) and growth outcomes at 2 years to assess if there were specific age periods of susceptibility .
Because of the near ubiquity of EAEC detection in these study sites , few factors were identified that were associated with EAEC detection in surveillance stools . Enrollment weight , exclusive breastfeeding , and recent macrolide use were the only protective factors in the multivariable analysis , and only the associations with the latter two had a substantial magnitude of effect ( Table 1 ) . Socioeconomic status ( WAMI ) was weakly protective , but the association was not statistically significant . Macrolide use in the past 15 days , but not cephalosporin use nor any other antibiotic use , was associated with a reduction in EAEC detection . However , macrolide use in the past 16–30 days was not protective ( RR: 0 . 94 , 95% CI: 0 . 85 , 1 . 05 ) . This short-term only effect of macrolide use was consistent across all sites and ages . Adjusting for age , site , and their interaction , EAEC was not associated with diarrhea and was found significantly more often in surveillance stools compared to diarrheal stools ( RR: 0 . 86 , 95% CI: 0 . 82 , 0 . 90 ) . This association remained when adjusting for recent antibiotic use and specifically macrolide use , as well as if restricted to only those children with no antibiotic use in the past 30 days . Similarly , presence of EAEC in stools was not associated with persistent diarrhea ( duration of 14 days or more; RR: 0 . 93 , 95% CI: 0 . 73 , 1 . 18 ) compared to non-diarrheal stools . EAEC detection was associated with higher contemporary concentrations of MPO ( MPO 0 . 14 ln ( ng/mL ) , 95% CI: 0 . 11 , 0 . 18 higher in the presence of EAEC ) , a marker of intestinal inflammation , at all sites ( Table 2 ) . It was also associated with higher levels of ALA ( permeability ) and NEO ( intestinal inflammation ) overall , with some variation across sites . However , the magnitudes of these associations were very small ( 1 . 15 ng/mL difference in MPO ) compared to the range of observed concentrations in the study ( MPO interquartile range: 2 , 050–12 , 920 ng/mL ) . In addition , EAEC was not associated with AGP , a marker of systemic inflammation , nor the lactulose-mannitol ratio , a marker of intestinal permeability , measured during the same month as the stool collection . EAEC was associated with elevated MPO independently of Campylobacter , but their combined effect on MPO was less than additive when both pathogens were present . Detection of EAEC alone was associated with an adjusted 0 . 17 ( 95% CI: 0 . 13 , 0 . 21 ) higher ln ( MPO ) concentration , Campylobacter alone was associated with an adjusted 0 . 19 ( 95% CI: 0 . 15 , 0 . 24 ) higher concentration , and the detection of both pathogens was associated with an adjusted 0 . 27 ( 95% CI: 0 . 21 , 0 . 34 ) higher concentration . Detection of EAEC was not associated with short term differences in growth velocity in both the one and three months following each monthly stool collection overall or at any site ( Fig 2 ) . Furthermore , there was no evidence of an interaction between EAEC detection and MPO in the same stool ( p for interaction: 0 . 9 and 0 . 5 for 1-month WAZ and LAZ velocity , respectively ) ; concurrent detection of EAEC and a high level of MPO were also not associated with short-term WAZ and LAZ velocity . Over the course of the first two years of life , there was no difference at 2 years in WAZ ( overall difference: -0 . 05 , 95% CI: -0 . 18 , 0 . 08 ) associated with a linear increase in EAEC stool positivity ( Fig 3 ) . In contrast , more detections of EAEC were associated with significant decrements in LAZ ( Fig 3 ) . The difference in LAZ at 2 years of age between a child at the 90th percentile of EAEC stool positivity from 0–2 years ( 50% stools positive ) compared to a child at the 10th percentile for EAEC stool positivity ( 11% stools positive ) was -0 . 30 LAZ ( 95% CI: -0 . 44 , -0 . 16 ) . Among site-specific estimates , this association was greatest in Brazil ( LAZ difference at 2 years: -0 . 89 , 95% CI: -1 . 24 , -0 . 54 ) and South Africa ( LAZ difference at 2 years: -0 . 70 , 95% CI: -1 . 09 , -0 . 31 ) . There was evidence for an antagonistic interaction between high frequency of EAEC detection ( at least 50% of stools positive ) and high frequency of Campylobacter detection on the adjusted LAZ difference at two years , such that high detection of both pathogens was associated with a similar decrement in LAZ ( -0 . 29 , 95% CI: -0 . 74 , 0 . 15 ) as that for high detection of either pathogen alone ( EAEC: -0 . 38 , 95% CI: -0 . 54 , -0 . 22; Campylobacter: -0 . 29 , 95% CI: -0 . 43 , -0 . 14 ) . A high frequency of EAEC detection during only one of the periods 1–6 months , 7–12 months , and 15–24 months was not associated with LAZ decrements , whereas high frequency of detection in any two of the three time periods was associated with small non-significant length decrements , and high frequency of detection in all three time periods was associated with the largest length decrements ( Table 3 ) . There were no additional differences in growth between children who had at least one detection of EAEC in a diarrheal stool compared to children who did not after accounting for EAEC detection in surveillance stools ( Table 3 ) .
We identified widespread acquisition of EAEC within the first few months of life across diverse settings in South Asia , South America , and Africa . In all sites except Peru , EAEC was detected at least once by two years of age in more than 90% of enrolled children . Slightly lower detection of EAEC in Peru may be due to the relatively high rates of macrolide use observed at this site in MAL-ED [43] . A high prevalence of EAEC in children with and without diarrhea was also found in the seven-site Global Enteric Multicenter Study , a prospective matched case-control study of moderate-to-severe diarrhea [11] . There was no evidence in either study that EAEC was a major cause of diarrhea of any duration . Few risk factors for EAEC were identified in this analysis , and surprisingly , components of socioeconomic status and our index , the WAMI , were not consistently protective . Only exclusive breastfeeding , enrollment weight , and recent macrolide use were associated with reduced EAEC detections . Exclusive breastfeeding is protective against enteric infections through multiple pathways , including limits on environmental exposure through contaminated food and water and directly through antimicrobial factors like lactoferrin and antibodies present in breastmilk [44] . The percent days of exclusive breastfeeding accounts for temporary cessation and return to exclusivity , and the protective association of this construct emphasizes that the age of first stopping exclusivity may be less important than the practice of exclusive breastfeeding itself , which may occur in multiple episodes [45] . The association of EAEC infections with lower enrollment weight is consistent with the increased susceptibility of malnourished mice to EAEC infection compared to well-nourished mice [20] . Antimicrobial resistance is a common feature of EAEC [46–48] , and at least one EAEC-specific resistance island has been characterized [49] . This island does not contain resistance genes for macrolides , which may explain the protective association with macrolide use ( unlike either cephalosporin or any class of antibiotic use ) . The specificity of protection by macrolides may provide EAEC with a competitive advantage over other enteropathogens since non-macrolide antibiotic use was highly frequent at many of the MAL-ED sites [43] . Further characterization of the antimicrobial resistance of these isolates will be necessary to confirm this hypothesis . Because only recent macrolide use was protective against EAEC infections , clearance of EAEC may be incomplete or more likely , reinfection with EAEC occurs quickly . In addition , alterations of the microbiota by macrolides could increase susceptibility to later EAEC infections , as is evident in murine infections [21] . Therefore , antibiotic use to clear EAEC infections is likely not justified; however , increasing the duration of exclusive breastfeeding ( even if in separated episodes ) may delay the acquisition of these common , potentially inflammatory infections . EAEC detection was associated with markers of intestinal inflammation , most strongly with increased fecal MPO . While the magnitudes of the associations were small relative to the range of observed concentrations , the increase in average levels of fecal MPO associated with EAEC [0 . 17 ln ( ng/ml ) ] was comparable to that seen with Campylobacter infections [0 . 19 ln ( ng/ml ) ] , which is a recognized cause of inflammatory enteritis [50 , 51] . EAEC has been previously associated with markers of inflammation , specifically with lactoferrin [52] and the proinflammatory cytokines interleukin ( IL ) -1b [14] and IL-8 [13 , 14 , 53] . The relevance of elevated intestinal inflammation to potential systemic inflammation associated with EAEC is not clear; there was no evidence that EAEC was associated with elevated AGP , a marker of systemic inflammation , though we note AGP was tested less frequently in this study and could have captured highly acute responses that may not have been temporarily coincident with stool sampling . The association between EAEC and intestinal inflammation suggests a potential mechanism for the observed association between EAEC and growth . Intestinal inflammation [54] and specifically higher levels of fecal MPO [55–57] , have been associated with poor linear growth among children in Brazil , Bangladesh , and the Gambia . However , because the magnitudes of association with inflammatory biomarkers were very small , this pathway may not be a major contributor to the overall growth impact , or equally , the biomarkers measured may be suboptimal markers . EAEC was associated with substantial decrements in LAZ at two years of age , and the magnitude of this association was similar to that reported for Campylobacter in MAL-ED [40] . However , the effects were less than additive , such that a high frequency of detection of both pathogens was associated with similar decrements as those associated with either pathogen alone . In contrast , EAEC was not associated with WAZ . The lack of association of EAEC with short-term growth velocity of either weight or length and the fact that the greatest impact of EAEC occurred among children with the highest frequency of detection during the first 2 years of life suggest that repeated high rates of exposure to EAEC prolonged over many months is necessary for the manifestation of overall length decrements observed at two years of age . Continual carriage and/or re-infection with a pathogen that is ubiquitous in the environment may limit the possibility for catch-up growth resulting in consistent linear shortfalls in the longer-term . This analysis provides a comprehensive longitudinal assessment of EAEC infections in early life across diverse low-resource settings , drawing on a large number of stool collections , biomarker assessments , and repeated anthropometric measurements . The study was limited by the potentially suboptimal assessment of pathogenic EAEC since the virulence genes for EAEC are not well understood [58] , and there may have been differences in strain variability across sites . Our gene probes , aatA and aaiC , were chosen as characteristic plasmid and chromosomal traits of EAEC , respectively [59] , and may not be perfectly discriminating for pathogenic EAEC . Genetic probes generally associate with laboratory phenotypes , not necessarily clinical disease [49 , 60] . In a study of children in Mali , aatA and aaiC were not associated with diarrhea when considering presence of either gene alone or in combination [61] . Furthermore , EAEC is able to acquire additional virulence genes that could increase its pathogenicity , such as the acquisition of Stx2 phage ( a characteristic of enterohemorrhagic E . coli ) in a German outbreak of EAEC-associated gastroenteritis [62] . The potential inability to distinguish pathogenic versus non-pathogenic EAEC may contribute to the weak associations observed between EAEC , inflammatory biomarkers , and short-term growth velocity . In conclusion , we found that EAEC infections were very common in the eight MAL-ED sites over the first two years of life . While often acutely subclinical , repeated EAEC detections were associated with longer-term linear growth deficits . Further work is needed better quantify the contribution of intestinal inflammation caused by EAEC to impaired growth . Refining our understanding of virulence traits may further help elucidate mechanisms of pathogenesis as well as the potential for vaccine-mediated or other approaches to control these increasingly recognized enteric pathogens . Because these infections may cause lasting consequences in terms of environmental enteropathy and relate to child growth deficits , a better understanding of the mechanisms involved and relevant biomarkers are critical to developing targeted interventions to prevent these consequences for the world’s poorest children .
|
Enteroaggregative E . coli ( EAEC ) are pathogens that infect the intestine and can cause diarrhea . They are also commonly identified among young children in low-resource settings , who can carry the pathogen without symptomatic diarrhea . We examined the risk factors for EAEC infections and their associations with child health outcomes over the first two years of life in eight low-resource settings of the MAL-ED study . EAEC infections were detected using molecular methods in more than 30 , 000 stools collected from 2 , 092 children in the MAL-ED study . We identified risk factors for EAEC and estimated the associations of EAEC with diarrhea , markers of intestinal health , and child growth . Almost all children were infected with EAEC at least once by two years of age . Exclusive breastfeeding , higher enrollment weight , and recent macrolide antibiotic use were protective against these infections . Although not associated with diarrhea in these children , EAEC infections were associated with intestinal inflammation and reduced length at two years of age . EAEC may impact child development , even in the absence of diarrhea , by causing intestinal inflammation and impairing child growth .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"antimicrobials",
"neonatology",
"children",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"maternal",
"health",
"pathogens",
"drugs",
"immunology",
"sociology",
"microbiology",
"social",
"sciences",
"biomarkers",
"pediatrics",
"diarrhea",
"age",
"groups",
"women's",
"health",
"signs",
"and",
"symptoms",
"antibiotics",
"gastroenterology",
"and",
"hepatology",
"pharmacology",
"campylobacter",
"bacteria",
"bacterial",
"pathogens",
"families",
"inflammation",
"medical",
"microbiology",
"microbial",
"pathogens",
"immune",
"response",
"people",
"and",
"places",
"breast",
"feeding",
"biochemistry",
"diagnostic",
"medicine",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"social",
"research",
"organisms"
] |
2017
|
Epidemiology of enteroaggregative Escherichia coli infections and associated outcomes in the MAL-ED birth cohort
|
Pseudomonas aeruginosa is one of the most virulent and resistant non-fermenting Gram-negative pathogens in the clinic . Unfortunately , P . aeruginosa has acquired genes encoding metallo-β-lactamases ( MβLs ) , enzymes able to hydrolyze most β-lactam antibiotics . SPM-1 is an MβL produced only by P . aeruginosa , while other MβLs are found in different bacteria . Despite similar active sites , the resistance profile of MβLs towards β-lactams changes from one enzyme to the other . SPM-1 is unique among pathogen-associated MβLs in that it contains “atypical” second sphere residues ( S84 , G121 ) . Codon randomization on these positions and further selection of resistance-conferring mutants was performed . MICs , periplasmic enzymatic activity , Zn ( II ) requirements , and protein stability was assessed . Our results indicated that identity of second sphere residues modulates the substrate preferences and the resistance profile of SPM-1 expressed in P . aeruginosa . The second sphere residues found in wild type SPM-1 give rise to a substrate selectivity that is observed only in the periplasmic environment . These residues also allow SPM-1 to confer resistance in P . aeruginosa under Zn ( II ) -limiting conditions , such as those expected under infection . By optimizing the catalytic efficiency towards β-lactam antibiotics , the enzyme stability and the Zn ( II ) binding features , molecular evolution meets the specific needs of a pathogenic bacterial host by means of substitutions outside the active site .
β-lactam antibiotics ( penicillins , cephalosporins , monobactams and carbapenems ) are the most dependable and frequently employed chemotherapeutic agents for eradicating bacterial infections [1] . Their safety and efficacy as antimicrobial agents derives from their ability to selectively inhibit cell wall biosynthesis , provoking bacterial cell wall lysis [2] . The newest types of β-lactam antibiotics ( e . g . carbapenems ) and expanded-spectrum cephalosporins ( e . g . cefepime ) , evade most common mechanisms of resistance against these compounds [3] . These compounds are currently used as “last resort” drugs for treating multi-resistant gram-negative pathogens [1] , [3] . The major mechanism of resistance against β-lactam antibiotics is the production of bacterial β-lactamases which catalyze cleavage of the antibiotic β-lactam ring rendering an inactive derivative [2] . β-lactamases fall into four classes ( A–D ) . Classes A , C and D are serine-β-lactamases ( SβLs ) which employ an active-site serine to catalyze antibiotic hydrolysis , while metallo-β-lactamases ( MβLs ) , or class B β-lactamases , are metallo-enzymes requiring one or two zinc ions for their activity [4] . MβLs gained importance in the 1990s as the principal mechanism of resistance against carbapenems ( imipenem , meropenem ) [5] , [6] , [7] . MβLs degrade all classes of β-lactams except monobactams and , unlike most SβLs , these enzymes are not susceptible to therapeutic β-lactamase inhibitors . This fact , together with the facile dissemination of MβL genes among different clinical pathogens , relegates them as a serious clinical threat [5] , [7] . Indeed , outbreaks of pathogens producing NDM-1 , IMPs , VIMs or SPM-1 MβLs are increasingly common worldwide [8] . Atomic structures reveal that clinically relevant MβLs ( subclass B1 ) possess similar active sites: indeed , residues binding the essential Zn ( II ) ions ( first sphere residues ) are strictly conserved ( Figure 1A ) [6] , [7] . Despite being “broad-spectrum” enzymes , MβLs exhibit quite different substrate profiles , which cannot be correlated to different active site structures [9] . Many structural and mechanistic studies have focused on the analysis of active site residues and the role of active-site flanking loops to account for substrate recognition of MβLs [6] , [9] , [10] . However , the mechanism by which different B1 enzymes are tailored to hydrolyze some antibiotics better than others is not known . The fundamental question remains: how does protein evolution occur among MβLs that are found exclusively and adapted to a particular host ? This problem represents a central issue in linking molecular features to organismal behavior . In the clinic this notion may contribute to therapeutic failure . Pseudomonas aeruginosa is one of the most clinically important non-fermenting Gram-negative pathogens , being well known for its ability to acquire genes encoding resistance determinants , such as the acquired MβLs [11] , [12] . In addition , P . aeruginosa harbors a host of virulence factors . Of particular relevance , SPM-1 is an MβL produced only by P . aeruginosa , while other MβLs have been found in many different bacterial hosts [12] , [13] , [14] , [15] , [16] , [17] , [18] . At the present time , the blaSPM-1 gene is associated with a single clone ( SP/ST 277 ) of P . aeruginosa . This clone emerged relatively recently in South America . This unique dissemination suggests either: 1 ) the blaSPM gene came from another organism and has expanded in SP/ST 277 because of a fitness advantage; or that 2 ) this genetic determinant may have been optimized to meet the need of its native host . SPM-1 in P . aeruginosa is therefore a unique system to analyze the role of host-specific constraints in molecular evolution . The structure of SPM-1 has revealed unique features among pathogen-associated MβLs [19] . Spencer and coworkers have shown that clinically relevant B1 enzymes share a hydrogen bonding network spanning below the active site base , generally known as second sphere residues ( Figure 1 ) [19] . This network is disrupted in SPM-1 due to the presence of two atypical second sphere residues: S84 and G121 , which replace the conserved D84/R121 couple ( Figure 1B ) [9] . Here we examine the role of these positions ( located outside the enzyme active site “in the second sphere” ) and their impact on antibiotic resistance in the native bacterial host , P . aeruginosa . We report that this unique combination of residues is able to provide resistance to anti-pseudomonal β-lactams such as latest generation cephalosporins and carbapenems [11] , while sacrificing the catalytic efficiency against other β-lactams . Our findings reveal that second sphere residues are able to modulate the substrate specificity of MβLs according to the requirements of the bacterial host In addition , we show that these second sphere residues optimize the zinc binding affinity of SPM-1 in the bacterial periplasm , providing P . aeruginosa antibiotic resistance under zinc-limiting conditions , such as those prevalent during bacterial infection [20] , [21] .
E . coli is usually employed as a model bacterial host to compare the ability of the different MβLs to confer resistance , even for enzymes which are not found in Enterobacteriaceae [9] . We designed a system aimed to reproduce the native conditions of expression of the blaSPM-1 gene . The complete blaSPM-1 transcriptional unit from the clinical strain P . aeruginosa 48-1997A [15] ( i . e . , including the natural promoter , the leader peptide for periplasmic location , the mature protein and the transcriptional terminator ) was amplified and subcloned into the broad-spectrum vector pBBR1-MCS5 [22] , replicative in P . aeruginosa PAO ( Figure 2A ) . P . aeruginosa PAO cells transformed with this vector ( pΔEP-SPM-1 ) were able to express SPM-1 , export and process it properly to the periplasmic space . Western blot analysis showed two SPM-1 forms of 30 . 6 and 27 . 5 kDa in whole cell extracts , corresponding to the precursor and mature species respectively [13] . Instead , the periplasmic fraction contained only the mature form of the enzyme ( Figure 2B , C ) . Accordingly , the transformed cells were resistant to imipenem . In order to assess the “flexibility” of positions 84 and 121 in accommodating residues different from the native ones ( S84 and G121 ) , codons 84 and 121 were individually randomized in blaSPM-1 by overlap-extension PCR [23] , [24] . The amplifications were targeted to the mature blaSPM-1 coding sequence , and then subcloned into the screening vector , so as to avoid undesired mutations in promoter and terminator sequences . In addition , codons 84 and 121 were randomized together looking for possible synergistic effects between these positions . Single-codon random libraries gave rise to 103–104 transformants , while the double-codon mutant library elicited >3×104 transformants . According to Poisson distribution , the libraries obtained have a probability of harboring a mutant blaSPM-1 gene with a specific codon at position 84 or 121 ( or a specific combination of codons ) >99% [23] . Sequences of ten randomly selected mutants from each library indicated no obvious bias . Active mutants were selected by examining the ability of the different libraries to confer resistance toward different types of β-lactam antibiotics in P . aeruginosa PAO ( Figure 3 ) . Paper discs embedded with different antibiotics were applied onto LB-Gm agar plates with P . aeruginosa PAO transformed with the randomized libraries . We employed a penicillin ( piperacillin ) , a third-generation cephalosporin ( ceftazidime ) , a cephamycin ( cefoxitin ) , and a carbapenem ( imipenem ) . Twenty bacterial clones exhibiting resistance ( i . e . , located within the halos ) were isolated for each library and for each tested antibiotic ( a total of 240 clones ) . Plasmids were extracted and blaSPM-1 was sequenced in each clone . In total , 16 different variants ( wild type SPM-1 , 10 single mutants and 5 double mutants ) were isolated . As expected , wild type clones ( with residues S84 and G121 ) were selected in all cases . Mutants G121A , S84N and S84N/G121S were also selected against all tested antibiotics ( Figure 3 ) . On the other hand , some substitutions were isolated depending on the screening antibiotic , implying that positions 84 and 121 modulate the substrate profile of the enzyme . Surprisingly , none of the selected mutants carried mutations G121R or G121H ( prevalent in B1 and B3 enzymes , respectively ) . Substitutions at position 84 , instead , displayed typical residues from B1 ( S84D ) , B2 ( S84G ) and B3 enzymes ( S84N ) , among others [9] . The S84D/G121S combination is present in the B1 enzyme IMP-1 , closely related to SPM-1 [13] . We then analyzed the resistance profile of the libraries . MIC values for P . aeruginosa cells expressing each of the selected SPM-1 mutants were determined against different antibiotics . Cefepime ( an antipseudomonal cephalosporin ) was added to the initial set of antibiotics . Expression of SPM-1 markedly increased resistance towards antipseudomonas drugs such as ceftazidime and cefepime ( 200–250 times ) , while for cefoxitin ( an antibiotic to which P . aeruginosa PAO is naturally resistant ) , the increase in MIC was only 7-fold ( Figure 4 ) . In general , single-codon variants S84G , S84N ( naturally present in B2 and B3 enzymes ) and G121A ( the most conservative substitution in this position ) display the highest MIC values after the wild type ( WT ) enzyme ( MIC values equal or up to 2-dilutions lower compared to WT SPM-1 ) . In fact , together with S84N/G121S , these mutants were the most ubiquitous in the antibiotic selection experiments . Synergistic effects between residues 84 and 121 are apparent when comparing double mutants vs . single mutants . For example , while S84G and G121S mutations were detrimental for resistance against piperacillin ( MIC values approximately half a dilution lower than for WT ) , the combination of both mutations generated an enzyme conferring higher levels of resistance than the wild type ( MIC value of 16 µg/ml for S84G/G121S vs . 10 µg/ml for WT SPM-1 ) ( Figure 4 ) . Surprisingly , the S84D/G121S combination , naturally occurring in IMP enzymes , was not among the most resistant mutants for any of the antibiotics assayed ( MIC values 2–3 dilutions lower compared to WT SPM-1 ) . Figure 4 summarizes our data showing that mutations had different impact in the bacterial resistance profile depending on the antibiotic ( selection criteria ) . Therefore , second sphere positions 84 and 121 are able to shape the resistance profile , and possibly the substrate specificity . In general , mutants conferred lower levels of resistance than wild type SPM-1 ( within a range of 5 dilutions in MIC values ) . Surprisingly , piperacillin is an exception , since four mutants outperform the wild type variant ( S84N , S84N/G121S , S84Q/G121S and S84G/G121S in up to half a dilution in MIC values ) . In the case of cefoxitin , the range of MIC values spanned by the different variants is smaller than for the rest of the tested antibiotics ( 3 dilutions vs . 4 or 5 dilutions ) . The impact of mutations on MIC values is more informative for the case of the antipseudomonas compounds ceftazidime and cefepime , where several single and double mutants provide levels of resistance comparable to the WT enzyme , with MIC values increasing by two orders of magnitude . For imipenem , G121A is the only mutant giving rise to a large MIC value ( 35 µg/ml vs . 48 µg/ml for WT SPM-1 ) . Enzymatic studies in vitro of MβLs have been useful to uncover structural and mechanistic aspects of these enzymes . However , these data rarely correlate with the in vivo behavior [9] . We attempted to correlate the MIC values with the hydrolytic profile of the different SPM-1 mutants assayed in periplasmic extracts of P . aeruginosa , i . e . , in an environment closer to in vivo conditions . The β-lactamase activity of SPM-1 mutants was assayed in periplasmic extracts of P . aeruginosa PAO ( in periplasma ) and normalized relative to the amount of enzyme present in the periplasm ( quantitated from Western blot gels ) . Given that SPM-1 is an efficient cephalosporinase in vitro , we focused on these substrates . We employed three substrates with antipseudomonal activity already used in the MIC experiments: ceftazidime , cefepime and the carbapenem drug imipenem , together with two first-generation cephalosporins devoid of antipseudomonal activity ( cephalexin and cephalothin ) . Hydrolysis rates in periplasma show a very good correlation with MIC values in the case of cefepime and imipenem ( Figure 5 ) . For these two substrates , only mutant G121A was competitive with the performance of wild type SPM-1 . Instead , in the case of ceftazidime , cephalotin and cephalexin , the hydrolytic performance of the wild type enzyme was surpassed by several mutants . Mutant S84G ( present in B2 enzymes ) , and to a lesser extent S84N ( present in B3 enzymes ) were the variants eliciting the best performance for first-generation cephalosporins . We conclude that the second coordination sphere modulates the substrate specificity in SPM-1 so that this enzyme is adapted to better hydrolyze the latest antipseudomonal antibiotics ( cefepime and imipenem ) while the catalytic performance against first-generation drugs is far from being optimized . In all cases , the activity of endogenous AmpC was negligible ( as revealed by the lack of activity in the “No SPM-1” control strains ) . Ceftazidime shows a different profile: albeit being an antipseudomonal drug , can be better hydrolyzed by several mutants than by native SPM-1 . However , P . aeruginosa has developed different resistance mechanisms against ceftazidime which do not affect cefepime and imipenem ( hyperproduction of endogenous AmpC , deregulation of efflux pumps or acquisition of ESBLs ) [11] . We therefore evaluated the role of SPM-1 in the resistance of the P . aeruginosa clinical strain against these three antibiotics . Disks embedded with ceftazidime , cefepime and imipenem were paired with disks containing dipicolinic acid ( DPA , an inhibitor of SPM-1 ) , on an agar plate inoculated with the clinical strain P . aeruginosa 48-1997A ( including its native blaSPM-1 gene ) and the control P . aeruginosa PAO expressing SPM-1 [25] . While halos of inhibition were similar for all antibiotics in the control strain , ceftazidime exhibited a reduced halo in the clinical strain ( Figure 6 ) . When DPA was added to whole cell extracts of the model strain , no residual activity was monitored . Instead , residual ceftazidime hydrolysis was present after addition of DPA to extracts from the clinical strain ( Figure 6 ) . We conclude that resistance against ceftazidime in the clinical strain is not exclusively due to expression of SPM-1 , and therefore this drug has not elicited a significant evolutionary pressure on this enzyme ( or a higher activity against this antibiotic was not necessary a part of the substrate spectrum in order to be acquired by P . aeruginosa ) . In fact , there is evidence of SPM-1 producing isolates of P . aeruginosa that express AmpC ( probably due to an hyperproduction phenotype ) and OXA-52 enzymes , supporting our hypothesis [26] . We postulate that SPM-1 in the bacterial host has been exposed to the evolutionary pressure of the administration of newest antibiotics such as cefepime and imipenem , thus adapting to better hydrolyze these compounds by changes in the second coordination sphere . At this point , it is intriguing that mutant G121A , providing high levels of resistance and hydrolysis rates in the periplasm , has not been selected during natural evolution . MβLs are exported to the bacterial periplasm as unfolded polypeptides [27] . Therefore , in the apo ( non-metallated ) form , metal site assembly ( giving rise to the active variants ) takes place in the periplasmic space [27] . We have recently shown that Zn ( II ) availability is limited in this compartment , and that MβLs with reduced Zn ( II ) binding capabilities are unable to confer resistance [28] . We determined the MIC values of P . aeruginosa cells with different SPM-1 mutants in media containing excess or limiting concentrations of Zn ( II ) against cefepime . MIC values were unaffected by Zn ( II ) supplementation in all cases . However , under metal deprivation conditions ( by adding the chelating agent DPA ) , strikingly distinct effects were observed for the different SPM-1 variants ( figure 7 ) . Mutant G121A , exhibiting a high specific activity in periplasma ( Figure 5 ) , was the most sensitive variant to metal deprivation ( MIC values are 64-fold lower in 1 mM DPA ) followed by variants S84K and G121D . Wild type and mutant G121S , on the other extreme , were almost unaffected by these conditions ( MICs values diminished in only one-dilution in 1 mM DPA ) . The bacterial growth was also unaffected by these conditions as assayed by MIC values for the strain without the SPM-1 expression system . The lack of effect of excess Zn ( II ) in bacterial resistance likely reflects the action of the CzcABC pump in P . aeruginosa which , by extrusion of excess Zn ( II ) into the extracellular medium , keeps constant the levels of periplasmic Zn ( II ) . Thus , the second coordination sphere exquisitely tunes the zinc binding ability of SPM-1 so that the enzyme has been evolved to provide resistance at metal limiting concentrations . As a result , the atypical S84/G121 combination allows SPM-1 to confer antibiotic resistance under these conditions . P . aeruginosa PAO periplasmic extracts revealed similar levels of periplasmic SPM-1 variants , with the exception of S84P and S84K mutants , which were undetectable . We analyzed the thermal stability of the mutants in periplasma , by studying the temperature dependence of ( 1 ) periplasmic β-lactamase activity or ( 2 ) SPM-1 solubility for each mutant . Hydrolysis rates were evaluated at 30°C after incubation at different temperatures . Instead , protein solubility was analyzed by Western blot quantitation of the levels of soluble SPM-1 variants after incubation at different temperatures . A plot correlating the hydrolytic activities ( or solubilities ) with the incubation temperature revealed in all cases a well-behaved sigmoidal behavior , that can be fit to obtain apparent Tm ( Tmapp ) values for each variant ( Figure S1 and Table 1 ) . Tmapp values determined from both strategies display an astonishingly good correlation . The four most stable periplasmic SPM-1 variants were G121S>S84/G121 ( WT ) ≈S84G/G121S≥S84G , while the combination S84D/G121S ( naturally found in IMP-1 ) was the least stable mutant together with G121D and , to a lesser extent , G121A . At this point we selected some representative mutants for further characterization ( S84D , S84G , G121A , G121S , G121D , G121N , S84D/G121S and wild type ) . Similar experiments were performed with the apo-derivatives of periplasmic SPM-1 variants , which were obtained by dialyzing the periplasmic fractions against EDTA and DPA metal-chelators , excess NaCl and finally metal-free reaction buffer . Tmapp values of the apo variants were estimated as before by β-lactam activity ( in this case supplementing reaction media with 2 µM Zn ( II ) ) and protein solubility ( Table 1 ) . Apo-derivatives exhibited a narrower range of Tmapp values , suggesting that the main differences observed in the stability of holo-derivatives are due to stabilization upon metal binding . Mutant G121S showed the largest metal-induced stabilization ( ∼30°C difference in Tmapp between holo and apo derivatives ) . Mutant S84D/G121S , on the other extreme , was marginally stabilized by metal binding . Mutants G121A and G121D precipitated during Zn ( II ) removal . Mutants showing larger differences in stabilities between the apo and holo form are expected to be those displaying large Zn ( II ) binding affinities . In agreement with this hypothesis , the variants exhibiting the highest metal-induced stabilization ( wild type SPM-1 , S84G , G121S and S84G/G121S ) were the least sensitive to metal deprivation ( Figure 7 ) . We explored the structural effects of these mutations by molecular dynamics ( MD ) simulations in wild type SPM-1 and three selected variants: G121A , G121S and S84D/G121S [19] . After 5 ns of MD simulations , a water molecule from the bulk solvent penetrates the active site of holo-wild type SPM-1 , occupying the vacant position between T115 and S84 ( Figure S2 ) , and reconstructing the hydrogen bond network present in all B1 enzymes [19] . In the apo forms , Zn1 ligands become mobile , mostly due to alterations in second sphere residues . The largest changes were observed in the cavity between residues T115 and S84 . In WT SPM-1 , the second sphere adopted two different conformations: ( a ) “holo-like” , in which the cavity was able to accommodate water molecules connecting residues T115 and S84 , and ( b ) “apo-like” , in which water molecules were excluded from this network , and T115 and S84 show a direct interaction ( Figures 8 and S3 ) . The holo and apo variants of G121S ( the mutant showing the highest metal-induced stabilization ) closely resemble the structure of the holo and apo forms of WT SPM-1 . Instead , in the case of the S84D/G121S , the second sphere residues are locked into an “apo-like” conformation , disfavoring metal binding . In the case of G121A , the Ala121 side chain avoids contraction of the cavity , locking the second sphere into the “holo-like” form . The low stability of apo G121A suggests that this conformation is not viable in the apo form .
SPM-1 from P . aeruginosa is unique among pathogen-associated MβLs in presenting the singular S84/G121 combination as second sphere residues , instead of the conserved D84/R121 couple [13] , [19] . Here we report a thorough study of the impact of mutations in these positions in the antibiotic resistance , specific enzymatic activity , metal binding features and protein stability . A major novelty in our approach is the extensive use of the native host , P . aeruginosa . This approach allowed us: ( 1 ) to perform a medium-throughput screening of activities and stabilities of a series of mutants with a high degree of reproducibility , ( 2 ) to correlate enzymatic activities reproducing the native conditions ( i . e . , within the cell ) during bacterial growth , which parallel the resistance profile and ( 3 ) to identify additional environmental factors which may not stem out from in vitro studies or by using E . coli as a model bacterial host [9] . In fact , MIC values of imipenem elicited by MβLs in E . coli are markedly lower to those determined in their natural hosts [9] , [29] . In the particular case of MβLs , these enzymes are active only when the Zn ( II ) availability in the periplasm allows proper metal uptake , therefore being much more dependent on the bacterial host than serine lactamases [27] , [28] . Expression of WT SPM-1 in P . aeruginosa selectively raises the MIC values against ceftazidime , cefepime and imipenem . These MIC values correlate with the in periplasma specific activities , in particular against cefepime and imipenem . The wild type variant , together with G121A , shows the highest specific activities against these two antibiotics . Instead , many single and double mutants in positions 84 and 121 outperformed wild type SPM-1 versus several antibiotics to which P . aeruginosa is intrinsically resistant . This remarkable substrate selectivity control by second sphere residues shows that the atypical S84/G121 combination present in SPM-1 has been fixed to provide resistance to anti-pseudomonal drugs , while sacrificing the catalytic efficiency against other antibiotics . Analysis of the resistance profile against cefepime by controlling the Zn ( II ) availability in the external medium reveals that G121A ( the only variant able to compete with wild type SPM-1 in terms of specific activity ) is extremely sensitive to metal deprivation . The fact that G121A is not a natural variant of SPM-1 despite the high resistance observed in metal-rich media suggests that evolutionary pressure has been exerted to select MβL variants capable of providing resistance in low Zn ( II ) environments . Native SPM-1 , instead , is able to confer resistance under conditions of Zn ( II ) deficiency . Indeed , during infection , the immune system produces large amounts of calprotectin , a host-defense protein that prevents bacterial colonization by chelating Mn ( II ) and Zn ( II ) [20] , [21] . Thus , optimization of the zinc binding capabilities is a crucial evolutionary trait for MβLs in their natural environment . This finding , together with a recent report highlighting the need of proper assembly of a dinuclear site in the active site of MβLs in the periplasm [28] , highlights the need to address the periplasmic bacterial mechanism of Zn ( II ) homeostasis and its role in antibiotic resistance , which have been largely overlooked . The role of second sphere residues in catalysis is an emerging issue in enzymology [30] . A hydrogen bond network connecting metal binding residues below the active site is meant to preserve the electrostatics and modulate the active site features . Directed evolution experiments on the B1 enzyme BcII enzyme revealed that mutations responsible of enhancing the lactamase activity were located in this hydrogen bond network [31] , [32] . As analyzed in detail by Spencer [19] and Oelshlaeger [33] , [34] in structural , modeling and mutagenesis studies , this network spans metal ligands His116 , Asp120 and Cys221 , and the second sphere residues 115 , 84 , 121 , 69 , 70 and 262 . The D84/R121 combination is the most commonly found in B1 enzymes [35] . Molecular dynamics simulations showed that water molecules can enter into the second sphere hydrogen bond network in SPM-1 . These calculations also support how changes in the second sphere can modulate the Zn ( II ) binding affinity , ultimately impacting in the resistance profile in limiting metal environments . Most acquired MβLs , such as enzymes from the IMP , VIM and NDM families present many allelic variants , in contrast to SPM-1 [36] . This difference could be due to the fact that SPM-1 , as we demonstrate here , is optimized to meet specific Pseudomonas requirements , in contrast to the other MβL genes , present in many different genera of bacteria . In a broader perspective , our approach allowed us to investigate how resistance determinants adapt to specific host requirements , linking fine details of the structural and biophysical features of the enzymes with bacterial fitness . More studies using this approach are required to account for the versatility and adaptability of MβLs to overcome the challenge imposed by new antibiotics .
Rabbits were housed and treated according to the policies of the Canadian Council on Animal Care guidelines on: antibody production ( http://www . ccac . ca/Documents/Standards/Guidelines/Antibody_production . pdf ) . All efforts were made to minimize suffering and the procedures were approved by the Bioethics Commission for the Management and Use of Laboratory Animals inside the Science and Technical Committee of the University of Rosario , under resolution number 490/2012 ( PICT-2008-N°0405 ) . Escherichia coli DH5α ( Gibco- BRL , Gaithersburg , MD , U . S . A . ) was used for construction of pΔEP-SPM-1 plasmid . Pseudomonas aeruginosa 48-1997A , originally identified in Brazil , was provided by M . Castanheira and M . Toleman [13] , and used as the source of blaSPM-1 . Laboratory strain P . aeruginosa PAO was used for transformation of mutant libraries , microbiological and biochemical studies . All strains were grown aerobically at 37°C in lysogeny broth ( LB ) medium supplemented with antibiotics when necessary . Molecular biology procedures were done according to Sambrook et al . Transcriptional unit of blaSPM-1 was PCR-amplified from a genomic preparation of P . aeruginosa 48-1997A using primers SPM-1-fw and SPM-1-rv ( Table 2 ) , both containing a BamHI restriction site , and subcloned into pBBR1-MCS5 plasmid [22] . The product was digested with XhoI and SmaI enzymes ( Promega ) to eliminate restriction sites EcoRI and PstI from the MCS of the plasmid . Extremes were made blunt by treatment with Klenow fragment ( Promega ) and then ligated with T4 DNA ligase ( Promega ) . Restriction sites EcoRI and PstI were introduced at the each edge of SPM-1 coding sequence by mutagenesis using primers EcoRI-fw , EcoRI-rv , PstI-fw and PstI-rv ( Table 2 ) . The resultant plasmid , pΔEP-SPM-1 , was introduced into P . aeruginosa PAO by electroporation as described [37] . All constructs and amplifications were verified by sequencing at the University of Maine ( Orono , USA ) . Codons corresponding to positions 84 and 121 ( BBL numbering [35] ) of SPM-1 were randomized individually by Overlap Extension PCR , as previously described [23] , [24] . Mutagenic primers were designed so as to contain random trinucleotides at the desired positions ( S84X-fw , S84X-rv , G121X-fw y G121X-rv , Table 2 ) , using pΔEP-SPM-1 as the template [23] , [24] . The products were subcloned into pΔEP-SPM-1 through EcoRI and PstI restriction sites ( thus avoiding unwanted mutations in promoter or terminator during PCR reactions ) , and the ligation mixtures electroporated in P . aeruginosa PAO . Electrocompentents from each mutant library ( S84X and G121X ) were spread in LB-agar plaques containing 30 µg/ml gentamicin , then collected and stored at −80°C . Library of double mutants S84X/G121X was constructed by submitting a plasmid preparation from S84X library to codon randomization of position 121 , in the same way as before . Selection of mutants capable of conferring some degree of resistance towards β-lactam antibiotics was done as follows . LB-agar plaques were inoculated with a bacterial culture ( O . D . 0 . 1 ) of each mutant library , and disks embedded with 10 µg imipenem , 30 µg ceftazidime , 1000 µg cefoxitin , or 10 µg piperacillin placed on top of the agar . Mutant clones growing in the area of the antibiotic gradients were picked and the sequence of blaSPM-1 further determined [23] . Production and/or resistance levels of SPM-1 in P . aeruginosa PAO pΔEP-SPM-1 or P . aeruginosa 48-1997A were assayed by pairing disks embedded with 1 . 5 mg dipicolinic acid ( DPA ) with disks containing 10 µg imipenem , 30 µg ceftazidime or 30 µg cefepime , onto LB-agar plaques inoculated with the corresponding bacterial culture ( O . D . 0 . 1 ) [25] , [38] . Minimal inhibitory concentrations ( MICs ) were determined on plaque by the dilution method [38] . P . aeruginosa PAO crude extracts were obtained through sonication of cells washed in Tris 10 mM , MgCl2 30 mM pH 7 . 3 followed by centrifugation at 4°C . Periplasmic preparations of P . aeruginosa PAO were obtained by shock with chloroform as previously described [39] . Contamination of periplasmic extracts with cytoplasmatic proteins was discarded by Western-blot with antibodies against cytoplasmatic DnaK [27] . Levels of periplasmic wild type SPM-1 and mutants were determined by Western-blot of periplasmic extracts with polyclonal antibodies against SPM-1 ( obtained after inoculating a rabbit with a mixture of recombinant SPM-1 and Freund's adjuvant ) and immunoglobulin G-alkaline phosphatase conjugate . Protein band intensities were quantified with the Gel-Pro Analyzer 4 . 0 software ( Exon-Intron , Inc . ) and normalized to a bacterial periplasmic protein arbitrarily chosen . Initial rates of hydrolysis were measured in a JASCO V550 spectrophotometer at 30°C in 300 µl of reaction media containing 300 µM of substrate and 10 µl of P . aeruginosa PAO periplasmic or crude extract in 10 mM Tris , 30 mM MgCl2 at pH 7 . 3 . For comparison , hydrolytic activities of periplasmic extracts were made relative to the amount of SPM-1 or mutant present in the extract , estimated by Western-blot anti-SPM-1 of the extracts normalized as before . In order to study the contribution of SPM-1 in whole β-lactam activity , crude extracts ( normalized in total protein concentration by Bradford assay [40] ) were incubated during 20 minutes at room temperature with and without addition of 25 mM DPA , and initial rates measured and compared . Aliquots from each periplasmic extract of P . aeruginosa PAO were incubated for 5 minutes at various temperatures in the range 30–90°C , and then placed on ice for ( a ) determining initial rates of hydrolysis against ceftazidime , or ( b ) determining the levels of soluble SPM-1 or mutants ( as before by Western-blot anti-SPM-1 of normalized extracts ) after centrifugation for 10 min at 10 , 000 rpm and 4°C . Activity curves or soluble protein fraction as a function of temperature was adjusted to the sigmoid curve f = y0+a/ ( 1+exp ( − ( x−x0 ) /b ) ) in Sigma Plot 9 . 0 program , with x0 the apparent melting temperature . In order to generate apo-derivatives of periplasmic SPM-1 and mutants , periplasmic fractions of P . aeruginosa PAO were dialyzed in duplicate against 500 mM EDTA , 500 mM DPA , 50 mM Tris pH 8 , then 2M NaCl , 50 mM Tris at pH 8 , and finally 10 mM Tris , pH 7 . 3 30 mM MgCl2 . The solutions were previously treated with chelating ion exchange resin ( Chelex 100 , Sigma-Aldrich ) and dialysis times were of 6 hours . All simulations were performed in AMBER [41] starting from the crystal structure of SPM-1 determined with resolution of 1 . 9 Å ( PDB code 2FHX ) [19] . As crystallization of SPM-1 was achieved with a vacant Zn2 site , the metal site structure of SPM-1 was reconstructed by aligning it to the geometry of the Zn2 site of the homologous enzyme B . cereus BcII ( PDB code 1BC2 ) [42] . In this way , a starting structure with a complete active site was obtained . Each simulation was performed using monomeric wild type SPM-1 , or mutant proteins G121S , G121A , S84D/G121S modified in silico . Furthermore , three crystallographic azide molecules were replaced by water molecules ( Wt1 , Wt2 and Wt3 ) in the cavities present at the base of SPM-1 active site . The systems were immersed in a box of water molecules TIP3P [43] and were simulated using periodic boundary conditions and Ewald sums for treating long-range electrostatic interactions [44] . The SHAKE algorithm was applied to all hydrogen-containing bonds [45] . This allowed us to use a time step of 2 fs for integration of Newton equations . Parm99 and TIP3P force fields implemented in AMBER were used to describe the protein and water , respectively [41] . The force field of the active site ( Zn , -OH , Asp , Cys and His ) was taken from the literature [46] . The temperature and pressure were controlled by the Berendsen thermostat and barostat respectively , as implemented in AMBER [41] . Cut-off values used for the van der Waals interactions were 10 Å . The systems were first minimized to optimize possible structural crashes and then slowly heated from 0 to 300 K under constant volume conditions , using a time step of 0 . 1 fs . Finally , a short simulation was conducted at a constant temperature of 300 K and under constant pressure of 1 bar , using a time step of 0 . 1 fs , to allow the systems reach a suitable density . These balanced structures were the starting points for the 10 ns of molecular dynamics simulations .
|
The presence of Zn ( II ) -containing metallo-β-lactamases ( MβLs ) that confer resistance to all penicillins , cephalosporins and carbapenems in Pseudomonas aeruginosa adds significantly to the threat of this pathogen in our health care system . SPM-1 is an MβLs widely distributed in South America and only found in P . aeruginosa . In common with all MβLs , the active site residues are highly conserved . In this work we asked the following question: how would substrate specificity evolve in SPM-1 if the active site residues are highly uniform and do not permit substitutions . To this end , we explored the role of two amino acids ( S84 and G121 ) that are outside the active site ( second sphere ) and are unique in the SPM-1 β-lactamase . We discovered that replacing these amino acids impacts resistance to cephalosporins and carbapenems and that this resistance profile depends on the enzymatic behavior and the availability of Zn ( II ) in the environment . This work demonstrates how protein evolution by means of subtle substitutions outside the active site meets the specific needs of a pathogenic bacterial host .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2014
|
Host-Specific Enzyme-Substrate Interactions in SPM-1 Metallo-β-Lactamase Are Modulated by Second Sphere Residues
|
The seven antigenically distinct serotypes of Clostridium botulinum neurotoxins , the causative agents of botulism , block the neurotransmitter release by specifically cleaving one of the three SNARE proteins and induce flaccid paralysis . The Centers for Disease Control and Prevention ( CDC ) has declared them as Category A biowarfare agents . The most potent among them , botulinum neurotoxin type A ( BoNT/A ) , cleaves its substrate synaptosome-associated protein of 25 kDa ( SNAP-25 ) . An efficient drug for botulism can be developed only with the knowledge of interactions between the substrate and enzyme at the active site . Here , we report the crystal structures of the catalytic domain of BoNT/A with its uncleavable SNAP-25 peptide 197QRATKM202 and its variant 197RRATKM202 to 1 . 5 Å and 1 . 6 Å , respectively . This is the first time the structure of an uncleavable substrate bound to an active botulinum neurotoxin is reported and it has helped in unequivocally defining S1 to S5′ sites . These substrate peptides make interactions with the enzyme predominantly by the residues from 160 , 200 , 250 and 370 loops . Most notably , the amino nitrogen and carbonyl oxygen of P1 residue ( Gln197 ) chelate the zinc ion and replace the nucleophilic water . The P1′-Arg198 , occupies the S1′ site formed by Arg363 , Thr220 , Asp370 , Thr215 , Ile161 , Phe163 and Phe194 . The S2′ subsite is formed by Arg363 , Asn368 and Asp370 , while S3′ subsite is formed by Tyr251 , Leu256 , Val258 , Tyr366 , Phe369 and Asn388 . P4′-Lys201 makes hydrogen bond with Gln162 . P5′-Met202 binds in the hydrophobic pocket formed by the residues from the 250 and 200 loop . Knowledge of interactions between the enzyme and substrate peptide from these complex structures should form the basis for design of potent inhibitors for this neurotoxin .
Clostridium botulinum neurotoxins ( CNTs ) are the most potent toxins known to humans since even one billionth of an ounce is fatal . Seven antigenically distinct botulinum neurotoxins are produced by the bacterium Clostridium botulinum and they share considerable sequence homology , and structural and functional similarity [1]–[3] . They are produced as inactive single chains of molecular mass 150 kDa and released as active dichains , a heavy chain ( HC , 100 kDa ) and a light chain ( LC , 50 kDa ) held together by an interchain disulfide bond [4]–[7] . HC comprising two distinct domains is responsible for binding to neuronal cells and translocation into cytosol . LC is the catalytic domain cleaving one of the three proteins forming the SNARE complex ( Soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptors ) required for docking and fusion of vesicles containing neurotransmitters to target cells [8]–[12] . The SNARE complex formation is prevented when any of the SNARE proteins is cleaved and accordingly blocks neurotransmitter release leading to flaccid paralysis and eventual death . Catalytic domains of BoNTs are zinc proteases and cleave SNARE proteins with stringent substrate specificity though they share significant sequence similarity . BoNT/A and BoNT/E cleave the synaptosomal-associated 25 kDa protein ( SNAP-25 ) while BoNT/B , /D , /F , and /G cleave the vesicle-associated membrane protein ( VAMP ) . BoNT/C is the only one that has dual substrate specificity , viz SNAP-25 and syntaxin [13] . The enhanced substrate specificity of CNTs is due to the recognition of substrates at remote sites called exosites in addition to the active site [14] . The potency and the ease with which these toxins can be produced make them potential bioweapons and bioterrorism agents . The Centers for Disease Control and Prevention ( CDC ) has declared them as Category A biowarfare agents . Currently , while experimental vaccines are available , only an equine trivalent antitoxin is available for post-exposure therapeutics with a limited therapeutic window [15] . One of the most effective ways a drug can act is by blocking the site where the substrate binds to toxin and accordingly the crystal structure of substrate-enzyme complex is essential to map out a strategy . Even though crystal structure of SNAP peptide ( 146–206 ) -inactive enzyme complex is available , it lacks interactions at the active site since the enzyme used was an inactive double mutant [14] . Here we present for the first time the structure of the substrate peptide , QRATKM containing the scissile peptide bond , bound to the active enzyme . This crystal structure reveals interesting features of complex formation which can help in designing efficient drug molecules to prevent or treat botulism . It is remarkable that this natural substrate peptide is not cleaved by the enzyme . In addition , we are also reporting the crystal structure of RRATKM , a variant of the substrate peptide , in complex with the enzyme . Though both are weak inhibitors , RRATKM is a better inhibitor than QRATKM .
Clostridium botulinum neurotoxin serotype A truncated light chain ( residues 1 to 424 ) , Balc424 , was expressed in E . coli and purified to homogeneity using size exclusion chromatography , as described previously [16] . The purified enzyme in 20 mM HEPES , 2 mM DTT , 200 mM NaCl , pH 7 . 4 was stored at −20°C until used . Amides of the peptides , QRATKM and RRATKM , were custom synthesized by Peptide 2 . 0 Inc . , Chantilly , VA20153 , USA . The stock solutions of the peptides were prepared with the above mentioned buffer . Balc424-QRATKM and Balc424-RRATKM complex crystals were grown using a range of protein/peptide molar ratio ( 1∶5 to 1∶30 ) . Both QRATKM and RRATKM complex crystals were grown by sitting drop vapor diffusion at room temperature . Briefly , 3 µl of the protein solution ( 15 mg/ml ) was mixed with an equal volume of a reservoir solution containing 20% PEG 8000 , 100 mM sodium cacodylate , pH 6 . 5 , 5% ethylene glycol and 200 mM ammonium sulfate . Thick plate-like crystals were obtained in five days and were flash frozen with liquid nitrogen using 20% ethylene glycol as cryoprotectant . The X-ray intensity data for both complex crystals were collected at X29 beamline of National Synchrotron Light Source ( NSLS ) using ADSC QUANTUM 315 detector . Balc424-QRATKM and Balc424-RRATKM complex crystals diffracted to 1 . 5 Å and 1 . 6 Å , respectively and belonged to the P21 space group with one molecule in the asymmetric unit ( Table 1 ) . All data were processed using the HKL2000 suite [17] . The structures of the complexes were determined by Fourier Synthesis using the acetate bound Balc424 ( Protein Data Bank id 3BWI ) as model followed by rigid-body refinement and simulated annealing . The composite omit map and the difference Fourier showed interpretable electron density for these hexapeptides . The best results were obtained with data collected from crystals grown with 1∶25 ( protein/peptide ) molar ratio . The peptide models were built with O [18] and further refined with CNS [19] until convergence . The final refinement statistics are shown in Table 1 . Models were validated with the Ramachandran plot using PROCHECK [20] . The proteolytic activity of balc424 was determined by HPLC using P[187–203] synthetic peptide as reported previously [21]; [22] . Briefly , balc424 enzyme ( 550 nM ) was incubated with the 17-mer peptide ( 1mM ) at 37°C for 30 min in the assay buffer ( 50 mM HEPES , 0 . 25 mM ZnCl2 , 5 . 0 mM DTT , pH 7 . 2 ) . IC50 values were determined by varying the concentration of inhibitors . The experimental data were analyzed using equation 1 , where I is the inhibitor concentration , y is the percent inhibition , with a slope factor ( s ) of 1 . 0 . ( 1 ) Coordinates and structure factors have been deposited to the Protein Data Bank . BALC424-QRATKM ( 3DDA ) and BALC424-RRATKM ( 3DDB ) . The SwissProt accession number for BoNT/A is P10845 .
The crystal structure has been determined to 1 . 5Å resolution . The model refined with R and R free of 18 . 4 and 20 . 1% , respectively . The final refined model contains 423 protease residues , 6 substrate residues , one sulfate and one zinc ions and 375 waters . More than 91% of residues are within the most allowed region of the Ramachandran plot . The electron density in the residual map ( Fo-Fc ) was well defined for the hexapeptide and QRATKM could be modeled unambiguously except for the side chains of K and M ( Figure 1A ) . It appears that K could take two rotamer positions . This is the first time an uncleavable substrate bound structure of an active botulinum neurotoxin has been reported and it has helped in unequivocally defining S1 to S5′ sites . Most notably , the amino nitrogen and carbonyl oxygen of P1 residue ( Gln197 ) chelate the zinc ion ( Figures 2 and 3 ) . The amino nitrogen has replaced the nucleophilic water as was shown earlier [16] . The crystal structure of Balc424 with a substrate analog RRATKM has been determined to 1 . 6Å resolution . The R and R free for the final refined model are 20 . 1 and 21 . 2% , respectively . The final refined model contains 423 residues of protease , 6 residues of substrate analog peptide , two sulfate ions , one zinc ion and 375 waters . More than 90% of residues are within the most allowed region of the Ramachandran plot . The substrate analog could be modeled unambiguously in the residual map ( Fo-Fc ) ( Figure 1B ) . Except for some minor variations of side chain orientations , the hexapetide RRATKM binds similar to the substrate peptide QRATKM ( Figures 2 , 3 and 4 ) . As in the case of QRATKM , the P1 ( Arg197 ) amino group and the carbonyl oxygen chelate the catalytic zinc and the nucleophilic water has been replaced . P1-P5′ residues occupy identical subsites as in QRATKM . This kind of interaction seems to be common with all peptide analog inhibitors [16] and probably plays a dominant role in inhibiting the catalytic activity . Though we have shown earlier that short tetrapeptides ( analogs of substrate ) are good inhibitors ( nM range ) , the hexapeptides are weak inhibitors [16] . The IC50 of QRATKM and RRATKM are 133 and 95 µM , respectively ( Figures 1E and 1F ) .
The side chain of P1-Q197 is exposed to the solvent region but makes a hydrogen bond with Glu164 OE1 ( Figures 3 and 4 ) . However , it is stabilized by various other interactions as well . N and O chelate zinc while O is also hydrogen bonded to Tyr366 OH which stabilizes the substrate and positions it for catalytic activity . Mutation of Tyr366 to Phe or Ala resulted in dramatic decrease in activity [23]; [24] . The amino nitrogen which has replaced the nucleophilic water is hydrogen bonded to Glu224 OE1 and OE2 ( the latter through a water molecule ) . It is known that variation in P1 does not affect the catalytic activity , probably due to most of the interactions being with the main chain atoms [25]–[28] . Mutation of Glu164 to Gln only had a marginal effect on the catalytic activity [23] . The only difference between QRATKM and RRATKM is at P1 residue . This was based on our previous experience with tetrapeptides [16] since the positive charge on Arg197 better complements the charge in the active site cavity . While P1-Gln197 makes a hydrogen bond with Glu164 , P1-Arg197 makes a salt bridge interaction with Glu164 thus making it more strongly bound ( Figures 3 and 4 ) . There are additional interactions with a sulfate ion nearby but this may be an artifact of crystallization . Other than this , residues from 198 to 202 in both structures superpose well except for minor variations in side chain orientations ( Figure 2B ) . The following discussion on subsites S1′ to S5′ applies equally for the both structures . P1′-Arg198 occupies the S1′ site formed by Arg363 , Thr220 , Asp370 , Thr215 , Ile161 and Phe194 . Phe163 , though slightly farther , also forms part of this subsite . The amino nitrogen and carbonyl oxygen of P1′ are hydrogen bonded to Phe163 O and Arg363 NH2 ( Figures 3 and 4 ) . These two interactions stabilize the substrate binding . When Arg363 is mutated to Leu or Ala , the activity decreases by 620 and ∼80 fold , respectively [23]; [24] . In addition , the guanidinium group of P1′ Arg198 forms salt bridges with Asp370 and P1′-Arg198 NE forms a hydrogen bond with Ile161 O . The salt bridge interaction between P1′-Arg198 and Asp370 is crucial since mutation of Asp370 reduced the catalytic activity by 250–600 fold [23]; [29] . The other major interaction is the stacking of guanidinium group of P1′-Arg198 with Phe194 ( Figure 3 ) . This stacking interaction also plays a major role in the activity since Balc424 Phe194Ala has ∼100 fold less activity [29] . Accordingly , both the electrostatic and hydrophobic interactions are crucial for catalytic activity . The S1′ site is fairly big and gives enough flexibility for Arg198 . In substrate analog tetrapeptide inhibitor complexes , it takes various rotamer positions [16] . In BoNT/A arginine hydroxamate complex structure , Arg hydroxamate occupies the S1′ site . But Zn is chelated by the carbonyl oxygen and the hydroxamate group . Also the direction of the peptide N to C is reversed [30] . S2′ site is formed by Arg363 , Asn368 and Asp370 , while S3′ subsite is formed by Tyr251 , Leu256 , Val258 , Tyr366 , Phe369 and Asn388 . P3′-Thr200 OG makes a hydrogen bond with Tyr251 OH . P4′-Lys201 is exposed to the solvent region . In the present crystal structure the side chain density for this residue is weak probably due to high thermal factors ( Figures 1C and 1D ) . However , one of the rotamer positions could form a hydrogen bond with Gln162 OE1 . This does not form a hydrogen bond in the complex structure of BoNT/A-SNAP-25 ( 146–206 ) ( PDB id = 1XTG ) . Instead Glu257 is close by , about 4 . 5Å . S5′ site is made of Tyr251 , Phe369 , Leu256 , Ser254 and Phe 423 . P5′-Met202 occupies this hydrophobic pocket ( Figure 4C ) . The crystal structure of SNAP-25 ( 146–206 ) peptide with an inactive double mutant ( Pdb id = 1XTG ) had identified the exosites as recognition sites distant from the active site [14] . However , the region of SNAP-25 peptide near the active site was disordered and could not be modeled very well . Comparison of the C-alpha position of the corresponding residues in the present structure shows that the C-alpha positions of these six residues are shifted . C-alphas of 197 , 198 , 199 , 200 , 201 and 202 are 4 . 34 , 3 . 84 , 3 . 55 , 3 . 13 , 5 . 69 and 6 . 12Å for the corresponding C-alphas in the present structures ( Figure 5A ) . In the absence of Tyr366 , SNAP25 residues near the active site move towards 250 loop increasing the distance from catalytic zinc . When the wild type light chain is used , the SNAP peptide is closer to the catalytic zinc and the 170 loop . This shift is probably due to either the disorder or the inactive mutant in 1XTG . One possibility is that since residues corresponding to α-exosites are missing in the short peptide , the whole peptide could have slid down . But this possibility is less likely since the β-exosite interaction is maintained in both the structures . Though the C-alpha atom of P5′ in the current structure and 1XTG are farther apart , the side chains occupy the same place . We conclude that this shift is due to the loss of interaction of SNAP-25 with Tyr366 which has been mutated to Phe in 1XTG . Because of this difference , P4′-Lys201 has potential interaction with Gln162 of the enzyme rather than Glu257 . The length of the anti-parallel β sheet formed near the 250 loop ( β-exosite ) in 1XTG ( 13 Å ) is almost double the length as in QRATKM ( 6 . 5 Å ) ( Figure 5A ) . Based on the above observations , the subsites as identified in this structure truly represent the substrate-enzyme complex interactions . Though the overall conformation of the enzyme in 1XTG and the current structure is very similar ( RMSD is ∼1 Å for 400 Cα atoms ) , loops 200 and 250 vary significantly ( Figure 5B ) . This conformational change may be either due to the recognition of α-exosites in 1XTG or just an artifact of crystal packing . In the current structure , loops 200 , 250 and 370 pack together tightly whereas in the 1XTG , 200 loop moved away . The C-alphas of Pro206 ( within 200 loop ) in 1XTG and QRATKM complex are ∼12 Å apart . Recently , the structure of a complex between the BoNT/A-LC and an inhibitory peptide N-Ac-CRATKML has been reported [31] . Though the direction of the polypeptide is the same , the inhibitory peptide ( N-Ac-CRATKML ) is shifted down by one residue compared to the substrate peptide QRATKM ( Figure 5C ) . This appears to be due to the effect of oxidation of Cys and the N-terminal blocking acetyl group . The cysteine is oxidized to sulfenic form . Both the sulfur and the OH group chelate the zinc ion unlike in QRATKM complex where the carbonyl oxygen and amino nitrogen of P1 residue chelate zinc ( Figure 5D ) . As a consequence , the acetyl group takes the C-alpha position of P1′ ( Arg198 ) and P1′ arginine moves to P2′ alanine's place . Moreover , P1 carbonyl oxygen interacts with Arg363 instead of Tyr366 . In QRATKM , P1' arginine forms salt bridge with Asp370 through guanidinium:carboxylate pair whereas in the N-Ac-CRATKML it is through a single NE and OD1 interaction . Interestingly , even though the C-alpha position has moved , Arg198 side chain takes a different rotamer position made possible by the size of the cavity and stays in the same pocket . In addition , P4' lysine interacts with Tyr366 while in the substrate peptide ( QRATKM ) it interacts with Glu162 . Hence the positioning of the inhibitory peptide ( N-Ac-CRATKML ) may not represent the substrate binding position as in QRATKM structure . In both cases the enzyme does not undergo significant conformational changes as it did in the structure of SNAP-25 ( 146–206 ) peptide complex [14] . N-Ac-CRATKML is a fairly good inhibitor ( Kι 1 . 9 µM ) [28] . But when the N terminal Cys is replaced with 2-mercapto-3-phenylpropionyl ( mpp ) the Ki improved to 300nM . Keeping this as a control various truncations were done [27] . Truncating the last three residues of the mpp derivative ( KML ) increased the Ki 100-fold while deletion of only the last two increased it only by ∼13-fold . The importance of Lys201 of the substrate may be attributed to the potential hydrogen bond the terminal side chain atom ( NZ ) makes with Gln162 . Mutation of Lys201 to Ala increased the Ki 10 fold suggesting that the Lys side chain interaction is crucial . When Thr200 of the substrate was mutated to Ala , Ki increased only marginally since the hydrogen bond with OG was lost . However , it is not clear from the present structure why Ala199Val will increase the Ki <10 fold . A simple modeling shows that the S2' subsite is big enough to accommodate a Val . Mutation of Arg198 to Lys increases Ki by more than 1000 fold . This is because both the salt bridge and stacking interactions are lost . It appears stacking may be important since ionic interaction between Lys201 and Asp370 is still possible . Though the present hexapeptide lacks Leu203 , truncation of this peptide had no effect on Ki . Saturation mutation studies based on the crystal structure of BoNT/A with SNAP-25 ( 146–206 ) has been used to define two regions , active site ( AS ) domain and binding site ( B ) domain in SNAP-25 [14]; [29] . SNAP-25 residues 193–202 form AS while residues 156–181 form B . Our hexapeptides form part of AS only . In the same work , two minimal length peptides have been tested for catalytic activity , D193EANQRATK201 ( SN/A1 ) and A195NQRATK201 ( SN/A2 ) ( the numbers correspond to our numbering scheme ) . While SN/A1 was cleavable by BoNT/A , SN/A2 was not , suggesting that the N terminal DEAN is required for cleavage . This probably explains why QRATKM which lacks DEAN was not cleaved in our case even though we used up to 1∶30 ratio of Balc424 to peptide . However , the major reason for the peptide not being cleaved is the amino group chelating zinc . Any extension beyond in the N terminal direction would change the character of this amino group and may not be able to chelate zinc . However , the earlier study used GST fusion protein to express the short peptide and might have some effect in binding to the enzyme . This is supported by the facts that I192DEANQRATKKMLGSG207 had 1/5th the activity compared to wild type [22] and the mutants A195C and N196C in the 17-mer SNAP-25 substrate peptide [28] insignificantly affected Km and kcat . The current structure confirms our earlier model for catalytic mechanism [16] . Glu224 acts as the general base in abstracting a proton from the nucleophilic water and also helps in shuttling protons to the leaving group . In addition , the roles of Arg363 and Tyr366 are to stabilize the substrate for proper positioning and orientation as the carbonyl oxygens of P1 and P1' are hydrogen bonded to Tyr366 and Arg363 . Tyr366 further stabilizes the oxyanion role of P1 carbonyl oxygen . Another molecular mechanism for BoNT/A recognition and cleavage of SNAP-25 has been proposed [29] . In that mechanism P5 ( Asp193 ) residue of SNAP25 is supposed to make the initial contact with the enzyme at the α-exosites by forming a salt bridge with Arg177 . This in turn aligns P4'-Lys201 to form a salt bridge with Glu257 . These interactions are supposed to broaden the active site and allow P1'-Arg198 to dock into the S1' site by both electrostatic and hydrophobic interactions . The current structure does not support such a mechanism . First , the substrate peptide is able to dock into S1' site even though the peptide lacks substrate residues upstream of P1 . Second , the S1' site of Balc424 with and without bound peptide is similar and there is no indication of any change in shape or size . Third , there is no possibility for Lys201 to make hydrogen bond contact with Glu257 . Accordingly , our crystallographic data show that Balc424 is well positioned for peptide binding and catalytic action without having to undergo a conformational change . However , the interaction of P4' with S4' substrate may be disrupted after cleavage and help the substrate to leave allowing uncleaved peptide to bind in its place . But there is no experimental or mutational evidence for that . Even though botulinum neurotoxins are declared category A biowarfare agents , effective drugs are yet to be developed . Antibody therapeutics is emerging but more than one antibody may be needed to contain the effect of a single serotype [32] . An equine antitoxin is also available for post exposure therapeutics . Small molecule inhibitors are being developed but the active site of botulinum neurotoxin is large and it would be better to have larger molecules or strongly binding peptidomimetic inhibitors to block the active site . The current structure where S1 to S5' sites have been mapped unequivocally will be a good starting point . This would at least give a serotype specific inhibitor that could be transformed into an effective drug for botulinum neurotoxin A . We have shown that the P1 residue could be changed to Arg without affecting the binding efficiency and in fact it has proved to be a better inhibitor since it complements the charge in that region . It is known that changing it to cysteine improves binding [27] . However , oxidation of Cys may cause a problem . The structural environment of P1 also suggests that an amino acid containing an aromatic ring may be better suited as it would improve stacking interactions . The hexapeptide could be extended by one residue at the N terminus . However , it might affect the chelation of zinc by P1 amino group . The requirement of P1' Arg is crucial for BoNT/A activity . However , changing it to Tyr will still keep the stacking interaction though the salt bridge would be lost . Arg198Ala abolishes the activity without affecting the Km value [33] . S2' site also suggests that it can tolerate bigger hydrophobic , aromatic residue . It is possible to introduce modifications in the peptides to bring rigidity , specificity and resistance from proteases . There are endless possibilities that can be tried with the information provided by this structure . Our biochemical assays with full length and truncated balc ( balc424 ) do not show much variation and hence the results are equally applicable to both . It is desirable to have a broad spectrum inhibitor to be effective across the serotypes and this structure will be a starting point .
|
Botulinum neurotoxins are the most poisonous substance to humans . The ease with which the bacteria can be grown , its potency and persistence have made it a potential bioterrorism agent , and accordingly , botulinum neurotoxin has been declared as Category A agent by the Centers of Disease Control and Prevention . Since it is both a potential bioweapon and a bioterrorism agent , it is imperative to develop counter measures and therapeutics for these neurotoxins , as none are available so far except experimental vaccines and an FDA-approved equine antitoxin . Our work presented here is an important milestone towards achieving this goal . The best antidote can be developed by blocking the active site of any enzyme . The crystal structures of substrate peptide–enzyme complex presented here map the interactions between the two and provide critical information for designing effective drugs against this toxin .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"biochemistry/structural",
"genomics",
"biochemistry/protein",
"folding",
"biochemistry/drug",
"discovery"
] |
2008
|
Substrate Binding Mode and Its Implication on Drug Design for Botulinum Neurotoxin A
|
Despite explosive growth in genomic datasets , the methods for studying epigenomic mechanisms of gene regulation remain primitive . Here we present a model-based approach to systematically analyze the epigenomic functions in modulating transcription factor-DNA binding . Based on the first principles of statistical mechanics , this model considers the interactions between epigenomic modifications and a cis-regulatory module , which contains multiple binding sites arranged in any configurations . We compiled a comprehensive epigenomic dataset in mouse embryonic stem ( mES ) cells , including DNA methylation ( MeDIP-seq and MRE-seq ) , DNA hydroxymethylation ( 5-hmC-seq ) , and histone modifications ( ChIP-seq ) . We discovered correlations of transcription factors ( TFs ) for specific combinations of epigenomic modifications , which we term epigenomic motifs . Epigenomic motifs explained why some TFs appeared to have different DNA binding motifs derived from in vivo ( ChIP-seq ) and in vitro experiments . Theoretical analyses suggested that the epigenome can modulate transcriptional noise and boost the cooperativity of weak TF binding sites . ChIP-seq data suggested that epigenomic boost of binding affinities in weak TF binding sites can function in mES cells . We showed in theory that the epigenome should suppress the TF binding differences on SNP-containing binding sites in two people . Using personal data , we identified strong associations between H3K4me2/H3K9ac and the degree of personal differences in NFκB binding in SNP-containing binding sites , which may explain why some SNPs introduce much smaller personal variations on TF binding than other SNPs . In summary , this model presents a powerful approach to analyze the functions of epigenomic modifications . This model was implemented into an open source program APEG ( Affinity Prediction by Epigenome and Genome , http://systemsbio . ucsd . edu/apeg ) .
Central to transcriptional regulation of gene expression is the regulation of the quantities of transcription factors ( TF ) bound to genomic regulatory sequences . The information used to quantitatively control TF-DNA binding is not only encoded in the genomic sequences , but likely is also embedded in the chemical modifications to the genomic sequences and the nearby histones [1] . The chemical modifications ( called epigenomic modifications ) include the addition of a methyl group or a hydroxymethyl group to the 5th carbon of cytosine ( 5-mC and 5-hmC ) and a number of posttranslational modifications to the histone proteins [2] . These modifications can alter the chromatin structure and function by changing the charge of the nucleosome or directly interacting with TFs [3] . In turn , TFs can tether DNA modification enzymes and histone modification enzymes to change the epigenomic modifications around the TF binding region . Hence , both the genomic sequences and the epigenetic modifications contribute to define the regional diversity of the regulatory genome . Less clear is how the genome and the epigenome jointly encode regulatory information , and how TFs interact with such regulatory information . The goal of this work is to model the three-way interactions among the TFs , the genomic sequence , and the epigenome , and thus allowing for predicting TF binding affinities in equilibrium states . Genome-wide distributions of TF-binding and epigenomic modifications can now be obtained by high-throughput sequencing methods [4] . The explosive growth of data urges the methodological developments that can achieve mechanistic understanding of gene regulation . In particular , quantitative models are needed to learn the regulatory rules implemented by epigenomic modifications . Two classes of methods were developed to study transcriptional regulation with different goals and mathematical foundations . The first class of methods aims at deriving regulator-target relationships or finding regulatory sequences and motifs . These methods were built upon statistical associations among sequence patterns , TF binding , and gene expression [5]–[11] . An advantage of this class of methods is that it is easy to incorporate new data types including epigenomic modifications . Indeed , using statistical enrichment and machine learning ideas , recent efforts have incorporated nucleosome positions [12] and epigenomic modifications to identify TFBSs [13] and regulatory genomic sequences [12] , [14]–[18] ( Table S1B ) . However , machine learning methods do not allow direct biophysical interpretation for their parameters , and therefore they do not make biological inferences as directly as the thermodynamic models ( see below ) . The second class of methods aims at deriving molecular mechanisms of TF-DNA interactions , using a thermodynamic framework ( reviewed in [19] ) . The intensity of TF-DNA binding was modeled as the equilibrium output of input sequences and TFs [20] , [21] . Partially due to a huge computational burden , this class of methods was originally restricted to analyze a few selected regulatory sequences in single-cell organisms , where a few simplified assumptions can be made [20]–[22] ( Model assumptions , Table S1A ) . These models were extended to analyze nucleosome positions [23] , [24] , gene expression in drosophila embryonic development [25]–[27] , and genome-wide TF binding data [28] . The latter development offered a unique advantage , which is the capability of gaining mechanistic understanding of TF-TF interaction and TF-DNA binding from genome-wide binding data . However , this class of models cannot easily take into account epigenomic modifications , which are argued to be more influential to TF-DNA binding than cooperative interactions between TFs [29] , [30] . Here we present a thermodynamic model that incorporates epigenomic modifications . This model can learn synergistic and antagonistic interactions between specific TFs and epigenomic modifications from genome-wide TF binding and epigenomic data . We were interested in a few open questions on the mechanisms of TF-DNA binding . First , to what extent does an epigenetic modification change the binding strength between a TF and a genomic sequence , which is composed of multiple strong and weak binding sites ? Second , is the epigenomic influence to TF-DNA binding invariant to the nucleotide composition of the genomic sequence ? Third , many TFs have preferred DNA recognition codes ( a . k . a . motifs ) ; are there TF-specific epigenomic recognition codes ? Fourth , does the epigenome modulate the variability ( noise ) of gene expression in an isogenic cell population ? Finally , what is the role of the epigenome in modulating individual variation of TF-binding among humans ? We used two complementary experimental systems to study the above questions . The first system is mouse embryonic stem ( mES ) cells . We recently assayed genome-wide distributions of 5-methylcytosine ( 5-mC ) , 5-hydroxymethylcytosine ( 5-hmC ) , histone variant H2A . Z , and acetylation of histone 3 lysine 27 ( H3K27ac ) [31] . We combined these data with published chromatin immunoprecipitation followed by sequencing ( ChIP-seq ) datasets of 5 other epigenomic modifications [2] , [32] , [33] and 9 TFs [34] from mES cells . This combined dataset allowed us to study TF-epigenome-DNA interactions relatively comprehensively . The second system is the white blood cells of seven people , which allowed us to explore individual differences in humans .
We developed a quantitative model for TF-DNA binding in a given epigenomic context . The goal of this model is to predict the binding intensity of a TF in any genomic region in any cell type , using the genomic sequence and the epigenomic modifications ( cell-type-specific data ) . This model incorporates four types of biophysical information: the active concentrations of the TFs ( learned from ChIP-seq data ) , the binding preferences of these TFs to DNA ( motif ) , the nucleotide composition of the genomic sequence , and the epigenomic modifications ( see Methods ) . Given input data including position-specific weight matrices ( PSWM ) , ChIP-seq derived TF binding sequences and binding intensities , and genome-wide distribution of epigenomic modifications , this model can learn cooperativity among TFBSs ( any number of strong and weak , homotypic and heterotypic TFBSs ) . More importantly , it can learn synergistic and antagonistic interactions between a specific TF and every assayed epigenomic modification . The learning process involves two steps ( Figure 1B ) . First , the model scans each epigenomic mark independently to identify those that interact with the transcription factor of the interest and modulate its binding affinities to genomic sequences . Second , these identified epigenomic marks are combined into one unified model to predict the binding affinity of any genomic regions . The model quantifies the improvements of predicted binding affinities by using the identified epigenomic marks ( Table S2 ) . Because this model operates at thermodynamic equilibrium , it does not make causal inferences about epigenetic and TF binding changes . We implemented this model into an open source program APEG ( Affinity Prediction by Epigenome and Genome , http://systemsbio . ucsd . edu/apeg ) . We recently published two types of 5-methylcytosine ( 5-mC ) data in E14 mES cells , using methylated DNA immunoprecipitation followed by sequencing ( MeDIP-seq ) and DNA digestion by methyl-sensitive restriction enzymes followed by sequencing ( MRE-seq ) [31] , . A total of 45 . 2 million reads were generated from MeDIP-seq , reflecting 1 , 495 , 114 methylated 200 bp genomic segments ( windows ) across the genome ( Text S1 ) . A total of 2 . 1 million MRE-seq reads were generated from a total of three restriction enzymes , covering 428 , 367 unmethylated windows . We used a selective chemical labeling method to pull down and sequence 5-hydroxymethylcytosine ( 5-hmC ) regions ( 5-hmC-seq ) [31] , [36] . A total of 58 million 5-hmC-seq sequence reads were generated , detecting 1 . 5 million 5-hmC marked windows in mES cells . We assayed the genomic distribution of histone variant H2A . Z and acetylation of Histone 3 Lysine 27 ( H3K27ac ) in E14 mES cells [31] . With 19 . 9 million ChIP-seq reads , 1 . 1 million 200 bp windows were found to contain H2A . Z . It has a small overlap with promoter regions ( 10 . 45% of H2A . Z marked windows ) , suggesting its substantial involvement in distal regulatory regions [37]–[39] . With a total of 19 . 8 million ChIP-seq reads , around 1 . 0 million 200 bp windows were marked by H3K27ac . It had a moderate overlap with promoter regions ( 16 . 66% ) , in line with the thought that it is primarily an enhancer mark [40] . Interestingly , H2A . Z and H3K27ac exhibited differential overlaps with 5-hmC marked windows ( 25 . 69% and 31 . 57% ) and 5-mC marked windows ( 17 . 69% and 12 . 98% ) , respectively . This suggests both H2A . Z and H3K27ac tend to overlap with 5-hmC more than with 5-mC ( both p-values<2 . 2e-16 , Chi-square test ) . Combining these data with 6 published ChIP-seq datasets [2] , [32] , [33] , we obtained genome-wide distributions of 9 epigenetic marks and 9 transcription factors in mES cells , which served as the dataset for our model-based analyses . Even though some epigenomic modifications are assumed to take some general roles in synergizing or antagonizing TF-DNA binding , little is known whether such epigenomic functions are specific to certain TFs or are general to every TF . To explore this question , we applied our new model to genome-wide distribution data of 9 TFs and 9 types of epigenomic modifications in mES cells ( assayed by ChIP-seq , MeDIP-seq , MRE-seq , and 5-hmC-seq ) . Thirty interactions between TFs and epigenomic modifications were identified , forming an interaction network ( Figure 2A , Table S2 ) . Here , “interaction” refers to the positive or negative correlation of an epigenetic modification and the binding between a TF and DNA . Among the 9 epigenetic modifications , H3K4me3 , H3K27ac , and 5-mC each interacts with a large number of TFs , forming hubs in the interaction network . Among the five epigenetic modifications that exhibited negative roles , only 5-mC represses the mES cell-specific regulators Oct4 , Sox2 , Nanog , and Stat3 . Compared to the hubs , H3K4me1 is more specific . It plays a positive role to the binding of Nanog , Sox2 and Stat3 . Even more specific are H2A . Z , 5-hmC , and H3K9me3 , which may negatively correlate with the binding of cMyc and nMyc . These data suggest that not all epigenomic modifications “uniformly” interact with every TF . Some epigenetic modifications may be associated with the binding of specific TFs . Considering TFs often have recognition preferences to certain short genomic sequences ( motifs ) , we hypothesized that there are TF-specific epigenomic motifs . By an epigenomic motif we refer to a specific combination of epigenetic modifications that is characteristic to the in vivo binding sites of a TF . To test this hypothesis , we estimated the association of every epigenetic modification and the binding of each TF , i . e . in Equation ( 5 ) . For each TF , we compiled the influences of epigenetic modifications as a column vector ( Figure 2B ) . These influences are not identical across TFs ( columns of Figure 2B ) . This suggests that analogous to DNA motifs , in vivo TF-DNA binding also have epigenomic motifs . A PSWM is used to describe DNA motifs [41] . We propose to use the vector of model-learned influences of the K epigenetic marks { , … , } to describe TF-specific epigenomic motifs , where A denotes the TF of our interest and K represents the total number of epigenetic marks . The epigenomic motifs can be used in combination with PSWMs to approximate the binding preferences of transcription factors in vivo . We hypothesized that the predictive power of TF binding intensities should be increased by incorporating the information of epigenomic motifs . In other words , if epigenomic motifs exist , they should help to better predict TF binding intensities than using DNA sequences alone . Three computational experiments were done to test this hypothesis . We chose the Nanog TF for these experiments , mostly because Nanog is an essential TF in ES cells and Nanog's DNA recognition motif is not well understood . In the first experiment , we removed the epigenomic data and fed our model with genomic sequences only . Without epigenomic data , our model degenerates into the STAP model [28] . STAP uses the sequences ( 500 bp ) and the TF-specific PSWM to predict TF binding affinities , taking into account all possible interactions among strong and weak TFBSs . To quantify the model's predictive power , we used the Pearson correlation between the ChIP-seq signals ( as observed binding intensities ) and the model-predicted binding intensities . Pearson correlations were 0 . 211 and 0 . 212 in the training and the testing datasets , respectively , providing a baseline predictive power ( Control-1 in red , Figure 3 ) . We then applied the model to test each epigenomic modification . H3K4me1 , H3K27ac and H3K4me3 largely increased the model's predictive power of Nanog binding intensities from the baseline ( red bars , Figure 3 ) . These three epigenomic marks were thus inferred as interacting with Nanog . To test the robustness of model inference , we changed the metric for quantifying prediction power into Spearman's rank correlation ( Figure S1 ) and varied window sizes ( Figure S2 ) . Neither of these changes affected the inferred interacting epigenomic marks . In the second experiment , we randomly shuffled the genomic positions of the observed epigenetic modification intensities , generating 200 permutated datasets . Feeding the permutated datasets to the model , we obtained a background distribution of predictive power ( Control-2 in red , Figure 3 ) . Using this background distribution , we identified three epigenetic modifications with which the model can significantly better predict TF binding intensities ( red bars with * in Figure 3 , permutation p-value = 0 ) . These three epigenetic modifications were identified as interacting with Nanog . This permutation experiment used the same number of model parameters and the same amount of data ( PSWM , sequence , and epigenetic data ) as the experiment using the original data . It rules out the possibility that the increased predictive power was due to increased model complexities . As the 3rd control experiment , we replaced the epigenetic modifications in mES cells with the epigenetic modifications of mouse adipose cells [42] and kept the other data intact . None of the four epigenetic modifications in mouse adipose cells significantly increased the predictive power of Nanog binding in mES cells ( green bars vs . Control-1 and Control-2 in red , Figure 3 ) , suggesting our learned TFBS-epigenomic interactions were cell-type specific . We asked to what extent the epigenome can predict TF binding without using the genomic sequences . Two control datasets were generated . First , each epigenomic mark was fed to our model without sequence data ( becomes invariant to in Equation ( 4 ) , solid red bars , Figure S3 ) . The enhancer and open chromatin marks H3K4me1 and H3K27ac were most strongly predictive of Nanog binding , followed by the promoter mark H3K4me3 . These data are consistent with the idea that open chromatin and hypersensitivity sites are predictive of transcription factor binding regions [18] . Interestingly , H2A . Z is the fourth epigenomic mark that is predictive of Nanog binding . The regulatory function of H2A . Z in mammalian cells remains controversial . While H2A . Z is generally thought as an active mark of transcription , it is negatively correlated with gene expression in a mES cell differentiation process [43] . The positive association of H2A . Z with Nanog binding suggests that H2A . Z may facilitate Nanog binding in undifferentiated mES cells . Second , we collected all ( 214 ) PSWMs from the JASPAR database as background motifs [44] . These background PSWMs were fed to the model with each epigenomic mark . The mean and standard deviation of the model predicted binding intensities from these background PSWMs were derived ( hollow red bars and error bars , Figure S3 ) . The predictive powers of these control datasets were compared to the predictive powers using both epigenomic and PSWM information ( blue bars , Figure S3 ) . The in vivo Nanog motif combined with epigenomic data ( solid blue bars ) increased the accuracy of predicted Nanog binding affinities than using epigenomic data alone ( red bars ) . More than 20% increases of predictive power were observed using Nanog motif and H3K4me1 or H3K27ac than using H3K4me1 or H3K27ac alone . Even larger increases were found in comparing in vivo Nanog motif ( solid blue bars ) with background PSWMs ( hollow red bars ) . The latter comparison used models with the same number of model parameters . It rules out the possibility that the increased predictive power was due to increased model complexities . The TF-DNA binding motifs derived from the enriched sequence patterns using in vitro binding assays do not always agree with the enriched motifs from in vivo binding assays [45] . Depending on the TFs , the differences in motifs derived from in vitro and in vivo experiments can be small [46] or large [28] . The causes of such differences are unknown . We hypothesized that some epigenomic modifications can synergize with DNA to produce a somewhat different binding preference of a TF than the binding preference of this TF to naked DNA . To test this hypothesis , we chose to further analyze the Nanog motifs derived in vitro [47] and in vivo [28] . We used the in vitro Nanog motif together with all epigenomic data to learn and predict in vivo binding affinities ( blue bars , Figure 3 ) and compared to the results from the in vivo motif ( red bars , Figure 3 ) . Without considering epigenomic data , the in vitro and in vivo motifs had similar predictive powers of ChIP-seq signals ( Control-1 in red vs . Control-1 in blue , Figure 3 ) . However , except for H3K27ac , adding epigenetic modifications to the in vitro motif did not increase the predictive power of Nanog binding . Even for H3K27ac , its contribution to predicting Nanog binding was much larger when combined with the in vivo motif than when combined with the in vitro motif ( red and blue H3K27ac bars , Figure 3 ) . This means the model failed to identify clear TFBS-epigenomic interactions with the in vitro Nanog motif , suggesting that the epigenomic motif is specific to the in vivo Nanog DNA binding motif . In several cases , including H3K4me3 , H3K27me3 , H3K36me3 , and 5-mC ( both MRE and MeDIP ) , feeding the model with epigenetic data together with the in vitro motif even slightly decreased its predictive power as compared to not using epigenetic data at all ( blue bars vs . Control-1 in blue , Figure 3 ) . This is because the model allowing for TFBS-epigenomic interactions is more complex than that without epigenetic data . However , there is no extra information added due to the lack of interaction between the in vitro motif and the epigenetic marks . These data explain the difference between the TF-DNA binding motifs derived in vivo and in vitro: although the Nanog sequence motifs derived in vitro and in vivo have similar binding affinities to the Nanog protein in vitro [28] , the in vivo motif predicted Nanog binding events with a higher sensitivity given the specificity ( Figure S4 ) . This suggests that only the in vivo motif may interact with epigenetic modifications . The in vivo binding intensities are determined by TFBS-epigenomic interactions and cannot be faithfully reproduced with the sequence motif ( either in vitro or in vivo ) alone . We asked how epigenomic modifications may theoretically modulate transcriptional noise [48] and the cooperativity of TFBSs . To address this question , we used constraint-based simulation studies [49] , with the constraints being the physical and empirical limits of TF concentrations and epigenomic states in eukaryotic cells ( Text S1 ) . Transcriptional noise is the variability of gene expression among cells in an isogenic population [48] , [50] , [51] . We asked whether the epigenome can modulate the level of transcriptional noise . We studied simple transcription systems with one TFBS , by examining the change in binding probability as a function of the concentration of the TF and the presence of epigenomic marks . Following the main assumption of thermodynamic models of gene expression , every cell in an isogenic cell population has the same probability of producing a transcript , denoted as p ( , where is defined in Equation ( 1 ) and is a constant ) . The expected number of transcripts is proportional to , therefore the variability of reflects transcriptional noise [23] . Without any epigenomic marks , the binding probability increased as the concentration of the TF increased , forming a sigmoid curve ( green curve , Figure 4A–B ) . In a transcriptional system with one strong TFBS , the binding probability should reach the half of the maximum binding probability when the TF concentration passes a low threshold [52] . With a weak TFBS , the half of maximum binding probability should be reached at a high threshold of the TF concentration . Because the range of TF concentrations is generally between 10 , 000 and 300 , 000 molecules per cell in fruit fly , mouse , and human cells ( reviewed by [29] ) , in our simulation of a strong TFBS , the half of the maximum binding probability was reached when the TF concentration reached 10 , 000 molecules per cell ( green curve , Figure 4A ) . In the other simulated system containing a weak TFBS , the half of the maximum binding was reached at the TF concentration of 300 , 000 molecules per cell ( green curve , Figure 4B ) . In the presence of an activation mark , the sigmoid curve of binding probabilities shifted to the left ( red curve , Figure 4A–B ) with no overlap to the original curve . Similarly , in the presence of a repression mark , the curve shifted to the right ( blue curve , Figure 4A–B ) . The dynamic range of TF binding probabilities , constrained by the range of TF concentrations , is a major indicator of transcriptional noise [23] . These constraint-based simulations provided a theoretical prediction that in the presence of a strong binding site , an activation mark decreases the dynamic range of binding probabilities and thus suppresses transcriptional noise , whereas a repression mark enhances transcriptional noise ( Figure 4A ) . However , in a transcriptional system with a single weak binding site , both activation and repression marks tend to suppress transcriptional noise ( Figure 4B ) . The key assumption to these predictions is that the half of total binding probability of a weak ( strong ) TFBS is reached at about the upper ( lower ) bound of the available concentrations of the TF . We asked whether the epigenome could modulate the cooperativity of adjacent TFBSs . To obtain a baseline ( no cooperativity ) for this analysis , in a simulation study , we fixed the TF concentration ( [A] in Equation ( 4 ) ) and compared the binding affinities between a strong TFBS and a weak TFBS in various epigenomic conditions . As expected , in the presence of an activation mark , the binding affinity increases as the intensity of this activation modification increases ( solid curves , Figure 5A ) , and the reverse is true in the presence of a repression mark ( dashed curves , Figure 5A ) . Moreover , an increase of epigenomic intensity produces a smaller difference in the binding affinities of the two TFBSs ( solid and dashed curves become closer as epigenomic intensity increases , Figure 5A ) . However , the binding affinity of a weak TFBS cannot surpass the affinity of a strong TFBS in any levels of an epigenomic modification ( neither the solid curves nor the dashed curves crossed , Figure 5A ) . In other words , when there is no cooperativity between TFBSs , under the same epigenomic condition , the order of binding strengths among different genomic sequences is fixed . Because TF concentration ( [A] ) is a multiplicative factor that is separate from the rest in the calculation of the binding affinity ( in Equation ( 4 ) ) , changing TF concentration would not change the contributions from other factors to the binding affinity ( ) . Thus , the analyses above hold for any TF concentrations . Next , we examined the cooperativity of adjacent TFBSs with simulations . With nearly no epigenomic modifications , a simulated genomic sequence containing two weak TFBSs exhibited a binding affinity larger than that of another sequence containing one weak TFBS ( dashed and solid blue curves at epigenomic intensity = 10−2 , Figure 5B ) , but smaller than that of a medium-strength TFBS and a strong TFBS ( green and red curves at epigenomic intensity = 10−2 , Figure 5B ) . As the intensity of an activation mark increased , the binding affinity of the two-weak-TFBS sequence first surpassed that of the medium-strength TFBS and later superseded the strong TFBS to become the sequence with the largest binding strength ( Figure 5B ) . This suggests that in the presence of the epigenome , the binding affinities of different genomic sequences may not always be monotonic . Considering that without cooperativity , the binding affinities of different sequences are strictly monotonic ( Figure 5A ) , these data suggest that epigenomic modifications are not only capable of increasing the binding affinity of each of the two weak TFBSs , but also can increase the cooperativity between the two TFBSs . Finally , we examined whether the binding affinity of the two weak TFBSs could surpass that of the medium-strength TFBS within the range of typical intensities of epigenomic modifications measured by ChIP-seq experiments . The dashed curve and the green curve crossed at the epigenomic intensity of 100 . 12 ( = 1 . 32 ) , corresponding to the enrichment ratio of e1 . 32 ( = 3 . 74 ) between the number of sequence reads in the input and the control samples . Because the enrichment ratio of these two numbers is typically between 1 and 40 [53] , the change of order of the binding affinities of these two simulated genomic sequences can happen in typical epigenomic conditions . With the theoretical understanding that epigenomic modifications can boost the cooperativity of weak TFBSs , we hypothesized that this is a general mechanism of quantitative regulation of gene expression . We explored this hypothesis with tri-methylation of Histone 3 Lysine 4 ( H3K4me3 ) and the transcription factor Oct4 , which is essential for maintaining undifferentiation [54] , [55] of mES cells ( Text S1 ) . Using Oct4 PSWM , we scanned all Oct4 binding regions , which were defined by the peaks in ChIP-seq data in mES cells [33] . We categorized the Oct4 TFBSs into two sets , strong TFBSs ( 2055 regions , Text S1 ) and weak TFBSs ( 1921 regions ) . The average H3K4me3 intensity on weak-TFBSs was larger than 150% of that on strong-TFBSs ( p-value<10−20 , Figure 5C ) . The largest difference of H3K4me3 intensities between the two sets appeared at the center of Oct4 binding regions ( Position = 0 , Figure 5C ) . This suggests that on Oct4 binding regions throughout the genome , H3K4me3 is more concentrated on those containing only weak sequence motifs . We ruled out promoters as a confounding factor to the association of strong H3K4me3 to weak TFBSs , because weak TFBSs do not preferentially locate in promoters ( Chi-square test p-value = 0 . 907 , Table S3 , Text S1 ) . We then asked if these weak-TFBS-only sequences could obtain a larger boost of binding affinity than the other sequences . Our simulation analysis suggested this was the case in theory ( Figure 5B ) . We now test it with the measured epigenomic and TF binding intensities in mES cells . We classified the ChIP-seq peaks into three sets , those only containing strong TFBSs , those containing both strong and weak TFBSs ( mixed ) , and those only containing weak TFBSs . We computed the change in Oct4 binding affinities on these sequence sets from not using H3K4me3 ChIP-seq data to using H3K4me3 ChIP-seq data . The weak-TFBS-only set exhibited a larger increase in binding affinities than the mixed set , which in turn had a larger increase than the strong-TFBS-only set ( Figures 5D , S5 ) . These data suggest that the endogenous levels of H3K4me3 in mES cells are sufficient to boost the binding affinity of adjacent weak TFBSs . Finally , had epigenomic boost of weak TFBSs happened in vivo , the model would be able to better reproduce in vivo binding intensities on weak TFBSs . To test this idea , we used DNA sequence and H3K4me3 to predict Oct4 binding regions and compared with ChIP-seq data . We quantified the improvements of the prediction accuracy between with and without considering H3K4me3 data . Applying the model with a stringent threshold on the predicted binding probability , the three sequence groups that harbor strong , mixed , and weak sites showed similar improvements on prediction accuracy ( strong-cutoff , Figure 5E ) . This indicates H3K4me3 helps to improve prediction , but does not specifically show it helps prediction on weak sites . Under two less stringent thresholds of the predicted binding probabilities , the model gained larger increases of prediction accuracy on weak sites and on mixed sites than on strong sites ( Medium-cutoff , Weak-cutoff , Figure 5E ) . These results are consistent with the idea that the model was able to predict the binding intensities more accurately by capturing the epigenomic boost of weak TFBSs . Besides H3K4me3 and Oct4 , several other epigenomic marks showed systematically stronger intensities near the weak TFBSs than near the strong TFBSs of other TFs ( Table S4 ) . Thus , epigenomic boost of the binding affinity of adjacent weak TFBSs is not only a theoretical possibility , but also can be a wide-spread regulatory mechanism . Genomic variations including single nucleotide polymorphisms ( SNPs ) can result in phenotypic variation . Still unknown is how epigenomes modulate the correlation of genotypes and phenotypes among humans . We chose TF binding intensities as a molecular phenotype to study this question . To study how epigenetic variation can interact with genomic variation , we did three between-individual comparisons across different ethnic groups . We first compared a European ( NIGMS catalog ID: GM12878 ) and a Nigerian ( GM18505 ) . We categorized NFκB binding regions with the TFBSs containing SNPs into two sets ( all analyses were done with homozygous SNPs , Text S1 ) . The first set had differences in NFκB binding intensities between these two individuals . This set was called Different Sequence Different Binding ( DSDB ) ( Figure S6A ) . The second SNP-containing set had similar NFκB binding levels in the two individuals , and were termed the Different Sequence No Difference in Binding ( DSNDB ) set . The first set ( DSDB ) was consistent with the theory that nucleotide changes in the TFBS should change the binding affinity of this TFBS; however , the second set ( DSNDB ) appeared to be inconsistent with such a theory . We hypothesized that the epigenetic marks on DSNDB stabilized the binding affinities of these binding sites . In other words , the epigenetic modifications on the TFBSs buffered sequence changes ( SNPs ) from changing binding intensities . Theoretically , the difference in binding affinities between two TFBSs is the largest without any epigenetic marks ( y-intercept , Figure 5A ) . When epigenetic modification intensities increase , the binding difference in the two TFBSs decreases ( from left to right , Figure 5A ) . This is true for any two TFBSs of the same TF . Thus , we have derived a theoretical mechanism for the epigenome to attenuate the TF binding differences on SNP-containing TFBSs in two individuals . We proceeded to examine whether the theoretical mechanism is relevant for transcription factor binding in humans . We first used our model to learn epigenetic marks that help to explain the binding intensities in all SNP-containing TFBSs ( Table S5 ) . Four epigenetic marks were identified by the model , which were H3K4me1/2 , H3K9ac , and H3K27ac ( Figure S7 ) . Among them , H3K4me2 and H3K9ac were identified as marks that better explain the binding intensities in DSNDB sites . If H3K4me2 and H3K9ac were used to attenuate binding differences between two people , there should be higher intensities of H3K4me2 and H3K9ac in DSNDB sites than in DSDB sites . Indeed , the intensities of H3K4me2 and H3K9ac were much higher in DSNDB sites than in DSDB sites ( p-values<10−20 , Figure 6 ) . To assess whether these results were specific to the chosen individuals in our analysis , we did two more comparisons . The second comparison was between a European ( GM12878 ) and a Nigerian ( GM19099 ) , and the third comparison was between a European descendant ( GM12878 ) and a Japanese ( GM18951 ) . Each comparison identified its own DSDB and DSNDB sites . However , all comparisons found significantly higher H3K4me2 and H3K9ac intensities in DSNDB sites than in DSDB sites ( Figure S8 ) . The NFκB binding intensities in DSDB and DSNDB of GM12878 had similar distributions , and therefore are unlikely to contribute to explain the differences of H3K4me2 and H3K9ac intensities in GM12878 ( Figure S6B–C ) . As a control , adding H3K36me3 data to the model did not increase the correlation of model predicted binding intensities to NFκB ChIP-seq data ( Figure S7 ) . Accordingly , the difference in H3K36me3 levels between DSDB and DSNDB sets was not clear and not consistent in these comparisons ( Figure S8 ) . Finally , we assessed whether inter-individual differences of PSWM matching scores were significantly different in DSDB and DSNDB regions . No significant differences were found in two pairs of individuals ( Figure S9 , S10 ) , ruling out the possibility that sequence-determined differences in binding energies were more pronounced in either of the two sequence sets . These data suggested a mechanistic explanation to the SNPs in TFBSs that do not produce between-individual differences in TF binding: epigenetic modifications on these TFBSs attenuated the binding differences . We note that not all factors relevant to gene regulation have been considered in this analysis . Other factors including DNase sensitivity and the binding of other TFs could play a role in buffering polymorphism in NFkB binding sites and therefore potential provide alternative explanations .
The overarching tenet of this work is obtaining mechanistic insights from high-throughput genomic data . Towards this goal , we forfeited commonly used “statistical enrichment” methods that look for large overlaps of two or more genomic features . Instead , we developed a biophysical model for the three-way interactions among the genomic sequence , the epigenetic modifications , and TF binding . The model is specified as a physical system , and every model parameter has a biophysical interpretation . This allows the analytical results obtained from this model to have mechanistic interpretations . Several epigenetic modifications were previously assumed to facilitate or hinder TF binding in a ubiquitous manner . For example , mono- , di- , and tri- methylations on histone lysine 4 ( H3K4me1/2/3 ) were thought to facilitate the binding of any TF . Our data suggested that some TFs tend to preferentially recognize TF-specific epigenomic codes . This implies that rather than ubiquitously synergize or antagonizing TF-DNA binding , some epigenetic marks can specifically interact with some TFs . This is conceivable because the maintenance of epigenetic marks often require histone or DNA modification enzymes to be brought to a genomic sequence by specific transcription factors [56] , [57] . In addition , epigenetic modifications are strongly associated with the three-dimensional ( 3D ) architectures of the local chromatin [58] . It is also conceivable that some TFs would have different binding preferences to the same DNA sequence but different 3D chromatin conformations . We showed that epigenetic modifications can boost the cooperativity of adjacent weak TFBSs . Thus , there is a functional advantage of coding a cis-regulatory sequence with a cluster of weak TFBSs rather than one strong binding site . The advantage is that the binding affinity of a cluster of weak TFBSs has a larger tunable range than a strong TFBS , in the presence of the epigenome . Thus , clusters of weak TFBSs offer the epigenome larger ‘controllability’ . This may explain why weak TFBSs tend to cluster in the mammalian genomes [59] . Consistently , we estimated that there were 2 . 3–4 . 6 weak Oct4 sites per ChIP-seq derived Oct4 binding region . Indeed , H3K4me3 was strongly enriched in Oct4 binding regions that only contained weak TFBSs . Moreover , H3K4me3 generated larger enhancements of binding affinities in the weak-TFBS-only binding regions than in other Oct4 binding regions . Thus , the ‘epigenomic boost’ of TFBS cooperativity can be a functional mechanism in mammalian cells . This provides an alternative view on the evolutionary origin of TFBS clusters , in which the presence of the epigenome was previously ignored [60] . A central question in personalized medicine is how genomic variation generates phenotypic variation . This is a challenging question because genomic variation was only partially correlated with TF-binding variation [61] . In particular , a set of SNPs in TFBSs does not introduce differences to TF binding as predicted by available TF-DNA binding models . Incorporating the epigenome into the TF-DNA binding model , we can now appreciate that some epigenetic marks can buffer genomic changes from generating changes in TF binding intensities . A case in point is that H3K4me2 and H3K9ac attenuate the personal variation of NFκB binding on SNP-containing binding sites in human lymphocytes . These results highlight the importance of considering the epigenome when analyzing the functional consequences of genomic variations . A limitation of the thermodynamic equilibrium model is that it does not make causal inferences . It does not differentiate whether an epigenomic motif promotes the binding of a TF , or the binding of a TF causes the buildup of an epigenomic motif . It is conceivable that TF binding and epigenomic motif can reinforce each other , in a sequence-dependent or sequence-independent manner . Recent cross-species comparisons reported larger evolutionary changes of TF binding regions [62] than epigenetically modified regions [43] . If we assume during evolution we should see larger changes in the effects than in the causes , then these data are compatible to the hypothesis that epigenetic factors could modulate the binding of specific TFs . Moreover , the model-identified epigenomic mark that have strong interaction with Oct4 binding is H3K4me3 , whose intensity is much larger in the Oct4 binding regions that contain only weak TFBSs than in the binding regions that contain strong TFBSs ( Figure 5C ) . If Oct4 binding had been the cause of H3K4 tri-methylation , we would expect H3K4me3 to be stronger on the regions containing strong Oct4 sites . Thus , at least in the case of Oct4 and H3K4me3 interaction , the data disfavor TF binding as the cause of this interaction .
First , a DNA sequence is associated with a physical state , which is defined by the combination of transcription factors bound to the sequence . When we consider one piece of genomic sequence a time , the physical state of a sequence can be regarded as the physical state of a cell . Second , TF-DNA binding has reached thermodynamic equilibrium , which implies the proportion of cells at each physical state does not change over time . Third , the binding affinity between a TF to any genomic location is a joint effect of multiple TFBSs in the “neighborhood” of this genomic location . Each TFBS has its own binding strength , and they may cooperate . Finally , the intensity of an epigenomic modification in this genomic neighborhood can influence TF binding . We model a genomic sequence ( S ) in a fixed epigenomic context as a physical system . Every TFBS in S can exist in one of the two physical states , occupied or not occupied by a TF . Thus , a sequence containing n TFBSs can exist in any of the total of 2n states ( Figure 1A shows the 23 states for a sequence containing 3 TFBSs ) . We use c to denote a state , and let C to denote all states . There is certain probability associated with every state of the system , denoted as P ( c ) . Such a probabilistic distribution is called a Boltzmann distribution [21] . From the perspective of a particular TF ( named A ) , the event that A is bound to sequence S is equivalent to the union of some of states of S . In the example in Figure 1 , the event ‘A is bound’ is the union of States 2 , 4 , 5 , 6 , 7 , and 8 . We call these states the occupied states ( O ) . Obviously , . The probability that A is bound to S is . We introduce the Boltzmann weight , , for every state c . is proportional to in the way that . Thus , the probability that A is bound to S is ( 1 ) We model the Boltzmann weight as follows . Two factors contribute to . The first factor is the binding affinity between the TF ( A ) and every TFBS , which is jointly determined by the TFBS and the epigenomic context . We denote this factor as qepi . The second factor is the cooperativity between TFBSs , denoted as . We formulate these thoughts as ( 2 ) where i and j are the indices of the TFBSs on S; and oi is the indicator of whether the ith TFBS is occupied ( , if occupied; , otherwise ) . This formulation implies that the state with no TFBSs bound ( for every i ) has a Boltzmann weight of 1 ( State 1 in Figure 1A ) . Suppose the ith and the jth TFBSs are bound by TFs A and B , respectively; is modeled as ( 3 ) We then model the binding affinity ( ) between TF A and the ith TFBS ( denoted by Si ) . Three factors can contribute: the TF concentration ( [A] ) , the preference of the TF to bind onto the binding site sequence Si ( denoted as K ( Si ) ) , and the epigenomic influence ( ) . These are modeled as ( 4 ) where k is an index for each type of epigenomic modification . Here K ( Si ) is the association constant of the binding site Si . We note that , where Scon is the consensus binding site and ΔE ( Si ) is the extra energy needed to bind onto a non-consensus sequence , which is correlated with the usual matching score between a TFBS ( Si ) and the PSWM of the TF . represents the influence of the kth epigenomic modification on the binding intensity on Si . We model the TFBS-specific epigenomic influence as follows . Let be the overall effect of the kth epigenomic modification to transcription factor A , ( 5 ) The TFBS-specific effect is a joint effect of the overall effect ( ) and the intensity of the kth epigenomic modification on Si ( denoted as ) . Taking the ChIP-seq data for a histone modification for example , is measured by the ratio of the number of sequencing reads between the experimental sample and the control sample . This study used the number of extended sequencing reads ( Text S1 ) falling on Si . We model the joint effect as ( 6 ) where σ is a threshold determining whether the measured intensity is beyond noise level . We note that implies either there is no detectable kth modification or the kth modification has no influence to the binding . Figure 1A illustrates how this model works for a sequence with three TFBSs and two partially overlapping epigenomic modifications . We call this model an epigenome-sensitive TF-DNA binding model . This model has two major applications . One is to predict the binding intensities of a TF throughout the genome in any cell type . The other application is to learn genomic-location-specific epigenomic influences on TF binding , i . e . . A third and relatively minor application is to learn the cooperativity between TFBSs in different epigenomic contexts . The required inputs are the genome sequence , the PSWM of the TF , and the epigenomic data . Epigenomic data are often generated by ChIP-seq , MeDIP-seq , and other sequencing based experiments . Standard analysis packages , including sequence mapping [63] and mapped reads postprocessing [42] can process each dataset into a genome-wide distribution of the intensity of an epigenomic modification . Our model takes such a distribution as an input through , the intensity of the kth epigenomic modification on Si . The model has two sets of models parameters , which are the cooperativity between TFBSs ( ωA , B ) and the influence of each epigenomic modification ( ) . To train these model parameters , four inputs are required . These include the genome sequence , the PSWM of the TF , the epigenomic data ( ChIP-seq and other forms ) , and the ChIP-seq data of the TF of interest . Let I ( A ) be the genome-wide distribution of binding intensities of transcription factor A . For example , if we segregate the human genome into 6 million 500 bp long windows , then I ( A ) is a vector of 6 million elements . Each element represents the ChIP-seq measured binding intensity in the corresponding window . Following previous notations , we use PA ( O ) in equation ( 1 ) to denote the model predicted binding probability of A in every window . We propose to learn the model parameters by maximizing the following target function ( 7 ) where corr ( . ) is the Pearson Correlation , and PA ( O ) is a function of and ωA , B . We implemented a maximization strategy to maximize f ( , ωA , B ) . The analytical form of P ( A ) can be explicitly expressed with a dynamic programming algorithm [28] . We maximize it by the Quasi-Newton Method ( a . k . a . BFGS algorithm ) provided in the GNU Scientific Library [25] , [64] . We start with random initial parameters and repeat it 500 times to avoid local minima . In applications where the cooperativity among TFBSs is not of interest , we propose to ignore the cooperativity term ( set ωA , B = 1 ) and only maximize with respect to . We identify an epigenomic modification k as associated with the binding of TF A when ≫1 ( positive ) or ≪1 ( negative ) . To test for the null hypothesis that = 1 , we shuffle the intensities of epigenomic modification k on the genome to obtain 200 random epigenomic profiles . We subsequently compute 200 values from the shuffled data and use them as the empirical null distribution . For each epigenomic modification k , we test = 1 using the empirical null distribution and reject the null hypothesis using a multiple-hypothesis-adjusted p-value [65] ( Figure 1B ) .
|
We developed a model-based approach to systematically analyze the epigenomic functions in modulating transcription factor-DNA binding . We postulated the existence of TF-specific epigenomic motifs , which could explain why some TFs appeared to have different DNA binding motifs derived from in vivo and in vitro experiments . The theoretical results suggested that the epigenome can modulate transcriptional noise and boost the cooperativity of weak TF binding sites . A preliminary analysis of the existing data suggested that epigenomic boost of binding affinities in weak TF binding sites could be a widespread regulatory mechanism in mES cells . Moreover , using personal data , we identified strong associations between H3K4me2/H3K9ac and the degree of individual differences in NFκB binding in SNP-containing binding sites , suggesting the theoretical mechanism for epigenome to attenuate the TF binding differences on SNP-containing binding sites in two individuals may contribute to link genomic variation to phenotypic variation . Thus , this model presents a powerful approach to analyze the functions of epigenomic modifications .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
Understanding Variation in Transcription Factor Binding by Modeling Transcription Factor Genome-Epigenome Interactions
|
Phenotypic differences of genetically identical cells under the same environmental conditions have been attributed to the inherent stochasticity of biochemical processes . Various mechanisms have been suggested , including the existence of alternative steady states in regulatory networks that are reached by means of stochastic fluctuations , long transient excursions from a stable state to an unstable excited state , and the switching on and off of a reaction network according to the availability of a constituent chemical species . Here we analyse a detailed stochastic kinetic model of two-component system signalling in bacteria , and show that alternative phenotypes emerge in the absence of these features . We perform a bifurcation analysis of deterministic reaction rate equations derived from the model , and find that they cannot reproduce the whole range of qualitative responses to external signals demonstrated by direct stochastic simulations . In particular , the mixed mode , where stochastic switching and a graded response are seen simultaneously , is absent . However , probabilistic and equation-free analyses of the stochastic model that calculate stationary states for the mean of an ensemble of stochastic trajectories reveal that slow transcription of either response regulator or histidine kinase leads to the coexistence of an approximate basal solution and a graded response that combine to produce the mixed mode , thus establishing its essential stochastic nature . The same techniques also show that stochasticity results in the observation of an all-or-none bistable response over a much wider range of external signals than would be expected on deterministic grounds . Thus we demonstrate the application of numerical equation-free methods to a detailed biochemical reaction network model , and show that it can provide new insight into the role of stochasticity in the emergence of phenotypic diversity .
Phenotypic heterogeneity in populations of genetically identical ( isogenic ) cells is one of the major discoveries resulting from a systems approach to molecular and cell biology . Application of single cell imaging techniques has demonstrated that individual cells in clonal populations may have very different phenotypes under the same environmental conditions [1] and that a pre-existing subpopulation of cells may survive a sudden environmental change that is lethal to the majority of cells , such as antibiotic treatment , thus gaining advantage [2] . These observations are particularly important in the context of survival strategies of bacterial pathogens . The phenotypic heterogeneity of isogenic bacterial populations has been implicated in the emergence of persistence and latent infection in Mycobacterium tuberculosis that makes this bacterium one of the most dangerous pathogens of mankind [3]–[5] . Phenotypic differences of genetically identical cells under the same environmental conditions have been attributed to the inherent stochasticity of biochemical processes [6] . According to theoretical predictions elementary chemical reactions involved in biochemical processes exhibit substantial stochastic fluctuations when low numbers of reactant molecules are involved within the small volume of a living cell . The existence of significant stochastic fluctuations in biochemical processes has been confirmed by numerous experiments including tracking of individual protein molecules in individual cells in gene expression processes [7] . The mechanism by which these fluctuations give rise to phenotypic diversity has been a subject of intensive study . In most cases phenotypic diversity has been attributed to stochastic fluctuations that result in switching between different stable states of the dynamical system occurring in a network that involves positive feedback loops [2] , [8]–[10] . Alternatively , a network may exhibit excitable dynamics , where fluctuations can lead to transient excursions from a single stable state to an unstable , but slowly decaying , excited state [11] , [12] . Yet another mechanism arises when a single stable state exists in the system , and the reaction network is effectively switched on and off according to the availability of one of the constituent chemical species [13] , [14] . Here we describe a novel situation , in which a monostable or bistable two-component system supports a persistent approximate basal solution , owing to stochastic delays in the transcription of either histidine kinase or response regulator genes . However , once a particular cell has reached a fully induced level of gene expression there is a negligible chance that it will revert to the basal state . Two-component signal transduction systems ( TCS ) are a very common mechanism by which bacteria sense external signals and induce the expression of genes that govern the response to environmental change . A particular environmental signal activates a specific membrane-bound histidine kinase ( HK ) , which in turn activates its partner response regulator ( RR ) via phosphoryl donation . The response regulator itself activates the transcription of multiple genes whose products enable the bacterium's adaptive response to the change it has sensed . A common experimental design is to introduce a reporter gene whose transcription is controlled by the response regulator , and to monitor the TCS output by measuring the number of reporter protein molecules produced by the reporter gene . We shall do the same in the numerical and analytical studies we present in this paper . We will later consider two scenarios: autoregulation of the RR gene , where RR activates its own transcription and so positive feedback is present , and the constitutively expressed RR gene , where activated transcription of RR is absent . It has already been shown that stochastic fluctuations in the expression of RR and HK genes lead to population heterogeneity with respect to the expression level of genes regulated by the TCS . Sureka et al . [3] used flow cytometry to show that the MprA/MprB TCS in Mycobacterium smegmatis leads to heterogeneous activation of the stringent response regulator Rel . that permits persistence to develop in Mycobacterium tuberculosis [4] , [5] . Sureka et al . complemented their experimental observations with numerical simulations of a stochastic kinetic model of the TCS , demonstrating that autoregulation of the RR results in bistable behaviour and that stochastic fluctuations in gene expression switch the system between the two stable states corresponding to two different phenotypes . Zhou et al . [7] had earlier used flow cytometry to measure gene expression in single Escherichia coli cells from a genetically identical population , in order to study cross-activation of the RR PhoB by noncognate HKs in the PhoR/PhoB TCS , and found a bimodal pattern of fluorescent protein reporter gene expression . Subsequently , Kierzek et al . [15] built the most comprehensive stochastic kinetic model of two-component system signalling published to date and used data of Zhou et al . to show that their model reproduces flow cytometry distributions of TCS-regulated fluorescent protein reporter gene expression . Further computer simulations demonstrated two response modes of the TCS leading to population heterogeneity . In the ‘all-or-none’ response that arises when the RR gene is positively autoregulated , the reporter gene is expressed either at fully induced or at basal level , and a change in the external signal strength results in a corresponding change in the fractions of cells expressing the gene at basal and fully induced level . Alternatively , population heterogeneity can be observed in a ‘mixed mode’ that occurs when the RR gene is constitutively expressed . In this response mode one population of cells expresses the gene at basal level , while in another cell population the gene is expressed at a level that depends on the signal strength . The mixed mode thus combines features of all-or-none and graded responses . In this work we use deterministic , probabilistic and equation-free methods to analyse the potential for simultaneous coexistence of different phenotypes in the Kierzek , Zhou and Wanner stochastic kinetic model of TCS signalling [15] ( hereafter KZW ) . The application of equation-free methods to biochemical reaction networks has typically focused on simple models of small networks [16] , [17] , though there have been some studies of larger scale networks [18]–[20] . Here we apply them for the first time , to the best of our knowledge , to a detailed model of signal transduction processes . Our results show that population heterogeneity can be generated by a molecular interaction network even when it is not multistationary . A deterministic bifurcation analysis of reaction rate equations derived from the KZW stochastic kinetic model shows that the mixed mode is absent in this framework . However , an equation-free analysis of the stochastic model , using the Gillespie algorithm with tau-leaping as a black-box time-stepper , in order to find stationary states for the mean of an ensemble of stochastic trajectories , reveals the long-term persistence of an approximate basal solution that combines with the graded response to produce the mixed mode . This confirms the results of a probabilistic analysis that establishes the essential stochastic nature of the mixed mode . The same techniques also show that stochasticity results in the observation of the all-or-none bistable response over a much wider range of external signals than would be expected on deterministic grounds . In summary , our work uses a detailed mechanistic model of the major signal transduction and gene regulation mechanism to show that multistationarity and positive feedback are not necessary for the emergence of phenotypic diversity and that deterministic bifurcation analysis is not always sufficient to explain phenotypic switching . In the Results section we first introduce the stochastic kinetic model that we shall be analysing , then we analyse the deterministic reaction rate equations that govern the chemical concentrations in the thermodynamic limit , and show that these do not permit a mixed-mode solution . In the following subsection we analyse the discrete stochastic system using equations for the expected ( probabilistic mean ) number of molecules of each chemical species present , and show that slow transcription of either or both of the histidine kinase or response regulator genes can lead to persistence of reporter gene expression at a level that is approximately basal when it would not be expected on deterministic grounds . In the final subsection of Results we show that equation-free methods can locate this unexpected basal expression solution and investigate its stability using only direct stochastic simulations . Thus we confirm the findings of our probabilistic analysis , and also demonstrate the potential of equation-free methods to shed light on stochastic effects in large complex systems where a probabilistic analysis is too difficult to perform . In the Discussion we summarise our findings and highlight their biological significance . The Methods section includes mathematical details of the probabilistic and equation-free analyses .
We base our stochastic kinetic model of the PhoBR TCS in E . coli on that of Kierzek , Zhou and Wanner [15] , summarised in Fig . 1 . ( A detailed representation of the model in Systems Biology Graphical Notation ( SBGN ) is given in [15] . ) We are interested in stochastic switching of reporter gene expression , and hence in the numbers of reporter protein molecules produced . The external signal is modelled as the ratio of the HK autophosphorylation to dephosphorylation rates . Dashed arrows on the diagram indicate activated transcription of the response regulator and reporter genes , modelled using the Shea-Ackers formalism [21] , where the reaction rate increases with the concentration of phosphorylated RR , saturating for large [RRP] at a level much higher than in the absence of RRP . As mentioned above , we will consider two cases: the autoregulated and the constitutively expressed RR gene . Transcription and translation of the response regulator , histidine kinase and reporter genes are modelled as pseudo-first-order reactions . The circle-headed arrows indicate HK/RR complexes in phosphate transfer processes , according to the Batchelor & Goulian model [22] . Included in our model but not shown in the diagram are dimer formation and dissociation and also reporter protein and mRNA degradation . KZW simulated the reaction network using the Gillespie algorithm [23] for direct stochastic simulation , and incorporating gene replication and cell division events . The Gillepsie algorithm updates the number of molecules of the chemical species , using the propensity functions , where is the probability that the reaction takes place in the time interval , and its associated stochiometric vector whose component is the change in caused by the reaction . The propensity functions for the reactions involved in the KZW model are given in Table 1 , where are the numbers of molecules of phosphorylated RR protein ( RRP ) , mRNA of RR ( mRNA-RR ) , RR protein ( RR ) , HK protein ( HK ) , phosphorylated HK dimer ( HK2P ) , complex of RR and phosphorylated HK dimer ( RR-HK2P ) , complex of phosphorylated RR and HK dimer ( RRP-HK2 ) , mRNA of reporter ( mRNA-Rep ) , reporter protein ( Rep ) , mRNA of HK ( mRNA-HK ) , phosphorylated RR dimer ( RR2P ) and HK dimer ( HK2 ) respectively , and are the corresponding concentrations . The correspondence between chemical species and model variables is also given in Table 2 for ease of reference . The parameters to given in Table 1 were chosen by KZW to accord with experimental data where available , or with validated models of prokaryotic gene expression or , in cases where it did not affect the qualitative results , they were chosen at will [15] . The concentration of RNA polymerase ( RNAP ) is fixed , at , in order to model transcription and translation as pseudo-first-order reactions , following KZW [15] , where is the cell volume and is the Avogadro constant and we set . The concentrations of the various degradation products mentioned in Table 1 do not influence the propensity functions and so we do not include them as variables in our model . The external signal is modelled as the ratio of the autophosphorylation to dephosphorylation rates for histidine kinase , , which we vary by keeping fixed and changing . In summary , the Gillepsie algorithm consists of randomly selecting the next reaction that occurs to be with probability proportional to , and randomly selecting the time , , until that next reaction takes place from an exponential distribution with rate parameter . The vector is updated according to the numbers of molecules created and consumed in reaction , and time is increased by [24] . Stepping forward in time in this way gives a single realisation of the system . Typically , many realisations are computed to give a fuller picture of the system behaviour . KZW started each realisation at at time , and performed 10 , 000 realisations , each of 20 , 000 s duration , for each parameter combination of interest . KZW were interested in two sets of comparisons: autoregulation of the RR gene versus constitutive expression , as discussed above , and fast versus slow transcription of HK . KZW chose an operating point for their system such that the mean steady state numbers of RR and HK protein molecules were 3800 and 25 respectively . The parameter values given in Table 1 are those for the autoregulated , slow transcription case . To simulate a constitutively expressed response regulator gene , we break the feedback loop by replacing the first two response regulator transcription reactions in Table 1 by the reaction , where prom-RR is the promoter region of the RR gene , with propensity function , where the rate constant is chosen to lead to the same system operating point in order to permit fair comparison with the autoregulated case . KZW found that a value of accomplished this [15] . In order to isolate the effect of variability in HK expression , KZW fixed the overall rate of transcription followed by translation to be . In the slow transcription , fast translation case the rate constants were and , while in the fast transcription , slow translation case these values were swapped . Slow transcription followed by fast translation produces HK in bursts , while fast transcription and slow translation leads to more continuous production [15] . With autoregulation of the RR gene and fast transcription of HK ( Fig . 2a ) KZW saw stochastic switching between the basal and fully induced levels of reporter gene expression - a so-called ‘all or none’ response . In other words , some trajectories showed very little reporter protein present at time , while some showed a large amount , and the number of reporter protein molecules produced during the productive trajectories did not seem to depend strongly on the external signal strength . The picture was similar with autoregulation and slow HK transcription , but there were fewer realisations at the activated level ( Fig . 2b ) . In the case of a constitutively expressed RR gene and fast HK transcription , there was no stochastic switching - a graded response was seen instead , where the number of reporter protein molecules produced increased with increasing signal strength ( Fig . 2c ) . An interesting novel case was found when the RR gene was constitutive , but transcription was slow , when stochastic switching and a graded response were seen simultaneously - a so-called ‘mixed mode’ ( Fig . 2d ) . It is the unexpected existence of this mixed mode that we seek to explain through our analyses below . In the thermodynamic limit where the cell volume and the numbers of molecules of each chemical species tend to infinity , but the concentration of each species remains constant [24] , the KZW model for the system containing an autoregulated RR gene can be reduced to the following set of deterministic reaction rate equations that describe mass-action kinetics for continuous real-valued concentrations: ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) ( 7 ) ( 8 ) ( 9 ) ( 10 ) ( 11 ) ( 12 ) If the RR gene is constitutively expressed , then equation ( 2 ) becomes ( 13 ) The deterministic rate constants are appropriately scaled versions [24] of those used in the stochastic kinetic model: for , for , for . In reality for this system some species remain low in number , fluctuating between zero and a small integer number . Thus we expect the deterministic continuous analysis based on these equations to give clues as to the system behaviour , but to fail to describe it adequately in some important respects . Note that equations ( 8 ) and ( 9 ) decouple from the rest of the system , being dependent only on the input value of , but not feeding back into the remaining equations through the values of and . Thus the reporter protein concentration , , is ultimately determined by that of the phosphorylated RR dimer , . We considered four sets of parameter values that gave every combination of fast and slow transcription with autoregulated and constitutively expressed RR gene . In each case we first found a stationary solution of the reaction rate equations for a particular value of the external signal ( ) numerically and then continued it over a range of external signals , where was varied , using the XPPAUT software package [25] to produce a deterministic bifurcation diagram . In the autoregulated case , the basal level of reporter gene expression is shown at zero external signal ( ) , where it corresponds to the following fixed point of equations ( 1 ) – ( 12 ) : ( 14 ) ( 15 ) ( 16 ) ( 17 ) ( 18 ) ( 19 ) ( 20 ) ( 21 ) More generally it can be shown that the fixed points of equations ( 1 ) – ( 12 ) are given by ( 22 ) ( 23 ) ( 24 ) ( 25 ) ( 26 ) ( 27 ) ( 28 ) ( 29 ) ( 30 ) ( 31 ) where and are the solutions of the nonlinear equations ( 32 ) ( 33 ) In the constitutive case , equations ( 16 ) and ( 23 ) become and equation ( 33 ) becomes ( 34 ) It is clear that , apart from the value of , the steady solutions depend on the rates of HK translation and transcription only through the product , which we have set to be constant . Thus the slow and fast transcription cases have the same fixed points in the deterministic framework . It turns out that these fixed points also have the same stability type over the range of external signals that we examined , and so the deterministic bifurcation diagrams are the same for fast and slow transcription . The autoregulated case ( Fig . 3a ) shows a classical bistable scenario , with a stable state corresponding to the basal level of expression of reporter protein , coexisting over a range of external signals ( ) , with a stable state corresponding to the activated level of expression . For only the basal expression solution exists , while for only the activated state is possible . The switch between these two states , gives a classical ‘all-or-none’ response: in a population of cells , for a given external signal , some will show activated expression of reporter protein and some will show only basal level expression . This case corresponds to Figs . 2a ( fast transcription ) and 2b ( slow transcription ) of the KZW results , where an ‘all-or-none’ response is indeed seen . On breaking the feedback loop to investigate the constitutively expressed RR gene , a graded response is seen ( Fig . 3b ) , where the amount of reporter protein produced rises steadily as the external signal is increased , and this solution is stable . Both Figs . 2c and d show KZW results using a constitutively expressed RR gene , but while they saw a graded response in the fast transcription case ( Fig . 2c ) they saw a mixed mode when transcription was slow ( Fig . 2d ) , where the basal level of reporter gene expression persists for at least 20 , 000 s in some cells even for quite large external signals . In our deterministic bifurcation analysis , this basal solution is absent and so there is no mixed mode . We deduce that the mixed mode results from stochasticity and/or discreteness . We want to find the approximate steady state of the discrete stochastic system . It does not have true fixed points , such that remains constant for all time . However , we can look for fixed points of the expected value , or mean , . As shown in Methods , evolves according to the differential equation ( 35 ) This is different from the reaction rate equations for the evolution of the vector of concentrations because in general for nonlinear [24] . Thus when the RR gene is autoregulated , the rates of change of the components of the mean are ( 36 ) ( 37 ) ( 38 ) ( 39 ) ( 40 ) ( 41 ) ( 42 ) ( 43 ) ( 44 ) ( 45 ) ( 46 ) ( 47 ) where , , and . If the RR gene is constitutively expressed , then equation ( 37 ) becomes ( 48 ) To look for a basal solution for these equations , we set for a zero external signal , and look for solutions such that . In Methods , we show that . Bearing this in mind , we find that the basal solution for the means satisfies , ( 49 ) ( 50 ) ( 51 ) ( 52 ) ( 53 ) ( 54 ) and that and correspond to fixed points ( if such can be found ) of the equations ( 55 ) ( 56 ) These last two equations are not in closed form , involving the higher-order moment , and so we cannot deduce from them whether a solution for and actually exists . If the basal solution does exist , then we see that , with the exception of the values of and it is the equivalent of the deterministic basal solution with the deterministic rate constants replaced by their stochastic equivalents . In order to understand the stochastic behaviour , the equations for the evolution of the mean are not sufficient . For a given realisation of the system , the must take non-negative integer values , and this discreteness turns out to be important in understanding the existence of basal solutions where they are not predicted by the mass-action or mean reaction rate equations . We have from equation ( 45 ) . In the case of slow transcription , where ( and where here and hereafter indicates the rounded value , with half integers being rounded upwards ) , the closest an individual trajectory can get to the fixed point of the mean is at . We can now look for fixed points of , which satisfy equations ( 36 ) – ( 44 ) , ( 46 ) and ( 47 ) , with the term involving being zero in equation ( 39 ) . In fact , in the slow transcription case we have and so if we can find a steady state for the remaining components of , we would expect it to persist over a timescale of approximately . Equations ( 43 ) and ( 44 ) show that the basal level of reporter protein production occurs when , in other words when no phosphorylated HK dimer ( HK2P ) is present , so we will look for a steady state solution that also has and call it . Thus we have , and we now look for values of the remaining components of that are consistent with this . We are no longer restricting the external signal , , to be zero . However , we find for all except for which . Thus an approximate steady state can be found that corresponds to a basal level of reporter protein production for arbitrary values of the external signal . For the parameter values used in our study , we find that in the autoregulated slow HK transcription case and in the constitutive slow transcription case . In both autoregulated and constitutive cases , these solutions are equivalent to the deterministic and mean basal solutions except for the values of , and . Thus we expect to see the basal level of reporter protein in a proportion of cells for all values of the external signal in the slow transcription case for both the constitutively expressed and autoregulated RR gene . In the constitutive case , this is the origin of the mixed mode , and in the autoregulated case it is why the basal solution is seen at unexpectedly high values of the external signal . Note from equation ( 36 ) that a requirement for the existence of an approximate discrete basal steady state with is that and equation ( 46 ) shows that this in turn requires that there be no RR-HK2P complex present ( ) . Equation ( 6 ) then implies we must have . This holds for the majority of trajectories over long times for both slow transcription cases , since slow transcription of HK means that the levels of mRNA-HK is typically zero ( ) , and when that is true we can find steady states where there is no HK protein ( ) or HK dimer ( ) and hence no phosphorylated HK dimer forms ( ) , as we have just shown . In the autoregulated cases , if we look for steady state solutions for the mean that have , then from equations ( 37 ) and ( 74 ) we have ( 57 ) and for our parameter values this gives , which indicates that the average number of mRNA-RR molecules present is very low and transcription of RR is slow . The closest an individual trajectory can come to this value is at , so we look for fixed points , with , that satisfy equations ( 36 ) and ( 38 ) – ( 47 ) with the left-hand sides equal to zero . If we can find such a solution , we would expect it to persist over timescales of about ( since at ) . Since ( no mMRNA-RR is present ) , and since for a basal solution we also have no RR2P ( ) or RRP ( ) and thus no RRP-HK2 ( ) , we find from equation ( 38 ) that it is consistent to have no response regulator protein ( ) and so again is satisfied . Thus in the autoregulated , fast HK transcription case we find the approximate basal solution such that , for , and , and correspond to fixed points of the equations ( 58 ) ( 59 ) ( 60 ) if they exist . ( Again these equations involve the second order moment , and so we cannot deduce from them the existence of a fixed point of the mean . ) For the parameter values of our study this gives for the autoregulated fast transcription case . Although the terms in prevent us from determining , and explicitly , we see that ( 61 ) ( 62 ) and hence , assuming that the solution does indeed exist , we must have ( 63 ) ( 64 ) where is chosen such that ( 65 ) holds . Since , must satisfy ( 66 ) For the parameter values just mentioned this gives , and from equations ( 63 ) and ( 64 ) we then see that and . Since , we must also have ( 67 ) This is automatically satisfied if , which is required if equations ( 63 ) and ( 64 ) are to have non-negative solutions for and . Note that in the autoregulated , slow HK transcription case , we can find an approximate basal solution that has both and equal to zero: in other words for and , with the growth rates of all components being zero except for and where the growth rates are and respectively . For the parameter values of our study we have . In the case of the constitutively expressed RR gene with fast HK transcription , no approximate basal steady state can be found . The rapid production of mRNA-HK ( ) in the fast transcription cases - a steady state of approximately molecules from equation ( 45 ) - ultimately leads to the production of phosphorylated HK dimer ( ) . The rate constant , , for the constitutively expressed RR gene is chosen to produce similar numbers of reporter protein molecules to those found in an activated cell in the autoregulated case . Thus , when phosphorylated RR dimer ( ) is scarce , RR transcription is much faster for the constitutively expressed than autoregulated gene . Equation ( 48 ) gives a steady state of approximately 10 mRNA-RR molecules in the constitutive cases , since , and hence RR protein ( ) is also present at high levels . The combination of both phosphorylated HK dimer and RR protein allows RR-HK2P complex ( ) and hence RR2P ( ) to form and ultimately leads to the presence of reporter protein ( ) at levels much higher than basal . Starting from the approximate basal solutions , we need RR protein ( ) , and prior to this mRNA of RR ( ) , and HK2P ( ) , and prior to this HK protein ( ) to form before reporter protein can be formed . This will happen only very rarely because either and are zero ( autoregulated cases ) or and are ( slow HK transcription cases ) or both ( autoregulated slow transcription case ) and the corresponding growth rates are tiny or zero , showing that the reactions involving these species are well-balanced at , and . Thus the approximate basal solution is expected to persist over long times for a significant proportion of trajectories , or equivalently in a significant proportion of cells . Only in the constitutive fast HK transcription case is there the required combination of nonzero ( and ) and ( and ) to cause the production of RR-HK2P complex ( ) and lead within a short time for the vast majority of trajectories ( or cells ) to the presence of reporter protein ( ) at levels above basal . This is the only case in which the basal solution is not observed for high external signals , as can be seen from Fig . 2c . Only a graded response is observed . The results show that slow transcription of either or both of the HK and RR genes can lead to the persistence of the basal solution where it would not be expected from analysis of the deterministic reaction rate equations . The discrepancy between the deterministic and discrete stochastic models arises from the fact that trajectories do not remain close to the basal level of expression for all time in the stochastic model when the basal solution is not a stable fixed point of the system . Rather they eventually approach the discrete stochastic equivalent of the steady-state solution found in the deterministic model . However , there is a delay before transcription of HK and RR is initiated during which a near zero level of expression is observed . HK transcription takes place at a ( stochastic ) rate to give mRNA-HK , which is then translated at a rate , where is the number of mRNA-HK molecules . For fixed , if the transcription rate constant is small , transcription occurs in bursts [15]: it is delayed for a long time in some realisations , followed by very rapid translation when the number of reporter protein molecules climbs up quite quickly towards its steady state value . Hence a basal level of expression is observed for a long time in some realisations of the discrete stochastic model . This is the origin of the mixed mode observed in the constitutive case ( Fig . 2d ) for slow HK transcription initiation . On the other hand if transcription is initiated rapidly , corresponding to large , the number of mRNA-HK molecules rises quickly and production of reporter protein occurs more steadily as long as RR is also being transcribed fast enough; thus trajectories depart from the basal solution earlier on average . The basal expression level is therefore not observed over long periods ( Fig . 2c ) . Note that the overall rate of transcription and translation of HK is the same in both cases , namely . If RR is transcribed slowly then this can also result in the basal expression level being observed over long periods , even if HK is transcribed rapidly , and this is why we see a persistent basal solution in the fast HK transcription autoregulated case . Bistable behaviour of stochastic origin has also been found in direct stochastic simulations of autoregulated gene expression [13] , [14] , where although mRNA transcription and translation are either not considered , or treated as a single lumped step , stochastic activation of the gene by binding of a protein dimer is required before gene expression can proceed . However , in that case , while dimer binding is sporadic , the remaining biochemical reactions in the network are comparatively fast , so that gene expression is effectively switched on or off by the presence or absence of the dimer and thus proceeds in bursts . At any given time some cells in a population would be switched off and so a basal expression state would be found when it was not expected on deterministic grounds , but the mechanism is different from the one we see here , where a given cell may persist in a basal state over a long period before transitioning to a higher level of reporter gene expression . The production of reporter protein at a level above basal , ultimately requires the simultaneous presence in the system of RR ( response regulator protein , ) and HK2P ( phosphorylated HK dimer , ) . This is much more likely to happen if both are present in significant numbers , as is forced to occur by the forms of the mRNA-HK and mRNA-RR growth rates in the constitutive fast HK transcription case , than if either RR or mRNA-HK appears only sporadically , which is true for the former if response regulator is initially scarce and the gene is autoregulated and the latter if HK transcription is slow . In these cases , we expect reporter protein production to continue at basal level over long times . As the system is stochastic there will always be trajectories that do lead to production of reporter protein at much higher levels , and indeed every trajectory would be expected to reach these levels if we were to wait long enough , because eventually there would be a stochastic fluctuation large enough to bring the trajectory into the basin of attraction of the induced expression solution . Since bacteria have a finite lifetime we would in practice observe reporter protein production at induced expression levels in a proportion of cells and at basal levels in the remainder . In the autoregulated slow HK transcription case , both values and are zero , so it is to be expected that after a given time a smaller fraction of cells in this case produces reporter protein at induced levels than in the autoregulated fast HK transcription or constitutively expressed slow HK transcription cases , and it can be seen from Fig . 2 that this is indeed the case . In the constitutively regulated slow HK transcription case , the expected value of ( response regulator protein ) is very high at , and so RR-HK2P complex ( ) and hence reporter protein ( ) will be formed rapidly if stochastic fluctuations lead to the presence of a few HK2P molecules ( ) . Thus after a fixed length of time , we expect a greater fraction of cells to show high levels of reporter protein in the constitutively regulated slow HK transcription case than in the fast HK transcription autoregulated case , where no more than about fifty HK2P molecules are present on average at steady state for the approximate basal solution ( ) , and so production of RR-HK2P complex ( ) will proceed much more slowly when occasional molecules of response regulator ( ) are formed . Again this confirms what is seen in Fig . 2 . In the previous subsection we analysed the equations for the time evolution of the mean directly in order to find the approximate basal solutions that give rise to the mixed mode and to the extended range of signals over which an all-or-none response can be seen . We were fortunate in being able to do this: many reaction networks would be too complicated to succumb to this approach . However , it is possible to use direct stochastic simulations to gain information about the existence and stability of steady states of the probabilistic mean . In this subsection we use equation-free techniques to confirm the existence of the approximate basal solutions and investigate their stability . This approach could be extended to complex reaction networks that cannot be analysed explicitly . So-called ‘equation-free’ methods ( see [26]–[28] and references therein ) are used to analyse the behaviour of dynamical systems that are either stochastic , or alternatively , deterministic of high dimension and with random initial conditions . The time evolution is obtained by a numerical time-stepping algorithm , and typically one is interested in characterising the asymptotic behaviour of the probability density functions of the associated state variables . Evolution equations for the probability distribution are often hard to write down in closed form , albeit their existence is guaranteed in most cases . However , ensembles of realisations of the dynamical system can be obtained by running the time stepper many times over for a given simulation time or time horizon , starting from a probability distribution of initial conditions . From these ensembles of realisations , moments ( typically the mean and sometimes also the variance ) of the probability distribution of state variables at the end time can be calculated . A key idea behind equation-free methods is that , if the high-order moments evolve much faster than ( are slaved to ) the low-order ones , there exists a closed evolution equation for the first few moments of the distribution . The method allows for the computation of steady states of , for example , the mean values of state variables , together with the corresponding Jacobian matrix that determines the stability eigenvalues for them , and so a bifurcation diagram can be constructed for these mean values . Thus all the powerful machinery of nonlinear dynamical systems can be brought to bear to explore systems for which explicit governing equations are not available . Typically the equation-free method also encompasses the identification of fast and slow state variables and the use of ‘coarse projective iteration’ to speed up the time-stepping in large systems . We have not implemented these aspects here , in the first case because we did not expect any separation of variables into fast and slow to remain valid over the entire range of parameter regimes that we need to investigate , since we are explicitly varying the timescales of interest in this problem , and in the second case because the use of modified tau-leaping in the Gillespie algorithm performs a similar role to coarse projective iteration . Equation-free methods have been demonstrated to work well for low-dimensional systems with tunable noise . They have also been used to examine stochastic simulations of ( bio ) chemical reaction networks in simple [16] , [17] , [29]–[31] and somewhat more complex cases [18]–[20] . Here we extend this work , by applying equation-free techniques to Gillespie algorithm simulations of a realistic biochemical reaction network of moderate complexity , which represents a significant computational challenge to the method . In order to capture the purely stochastic near-zero solutions involved in the mixed mode ( constitutive slow transcription case ) and the extension of the basal expression level to high external signals in the autoregulated cases , we use an equation-free method [32] in which the Gillespie algorithm is a black-box time-stepper . We begin by identifying microscopic and macroscopic variables for the system . The microscopic variables are contained in the vector , denoting the number of molecules of each species at time . The coarse variables of our problem are then defined as an ensemble average of over a large number , , of realisations of the Gillespie algorithm ( 68 ) where are the values of found in the realisations 1 to . A central role in the equation-free framework is played by the coarse time-stepper ( 69 ) The operator evolves the macroscopic state from time to time and , in general , is not available in closed form . However , it is possible to advance the coarse variables in time using independent microscopic runs of the Gillespie algorithm . The coarse time-stepper is then composed from these microscopic runs in three stages: lift , evolve and restrict [27] as described in Methods . Once the coarse time-stepper is defined , we can find steady states of the coarse evolution ( 69 ) by computing solutions to the equation ( 70 ) In our implementation , we find via Broyden's iterations: function evaluations consist in performing the lift-evolve-restrict steps mentioned above , whereas the Jacobian at points is determined numerically from the values of for various small perturbations of the mean . By choosing the time horizon , , appropriately we can pick up metastable solutions that persist , on average , for that length of time , but are not true steady states of the system . In practice , this turns out to be less straightforward than one might wish . The identification of a single fixed point requires hundreds of thousands to millions of realisations of the Gillespie algorithm ( owing to the use of a large ensemble and the requirement for several iterations of the algorithm before convergence ) , and is consequently very slow , even when the calculations are parallelised . The error tolerance that can be achieved depends on the number of realisations in the ensemble , and so there is a trade-off between accuracy of the solution detected , the time horizon required and the practical feasibility of performing the calculation . Nevertheless , this method confirmed the insights described in the previous subsection . In the equation-free root-finding algorithm , we use realisations and we set a relative tolerance of for Broyden's method . Finally , the time horizon varies between 30 s and 500 s; as pointed out in [28] , we expect the results to depend upon . Note that in determining we use the value of in each realisation that is computed at the last value of such that . We typically observe convergence of the Broyden's method within iterations , with the exception of a few points in the calculations of induced expression states where the solution jumps and the tolerance is met within or iterations . Since we are using a relative tolerance , our residuals never exceed , as the norm of our solutions is bounded by . Since the production of reporter protein , , is controlled by the number of phosphorylated RR dimers present , , in this subsection we use the value of , the mean value of , to illustrate our results . We first use the approximate basal solutions , , as an initial guess for the steady states at a very low value of the external signal in the constitutively expressed slow and fast transcription and autoregulated slow transcription cases , and use Broyden's method to find a nearby steady state . We then use this as a starting estimate of the solution at slightly higher external signal , converge once more to a nearby steady state , and in turn use this to find a solution at slightly higher external signal again . By this procedure of so-called ‘poor man's continuation’ we aim to trace out the dependence of the basal expression level of on the external signal . In the autoregulated fast transcription case , our initial guess at the lowest external signal level is . Fig . 4 shows that with the exception of the constitutively expressed fast transcription case , a metastable basal solution with persists at all values of the external signal between and for a time horizon of 300 s . When the time horizon is increased to in both slow transcription cases , and in the autoregulated fast transcription case , we start to see the loss of this persistent basal solution at medium to large external signals ( Figs . 4a , b and d ) . The profile departs from zero for some values of the external signal , whereas the ( underlying ) profile never does . We do not see a systematic variation of with external signal , because the level is still very low , and there is a certain variability in the numerical results that comes from using ensembles of stochastic trajectories . Thus , for example , no meaning should be attributed to the fact that the value of is zero in Fig . 4d for very high external signals , while it is nonzero for a range of signals below that: it is the fact that there are some nonzero values that is important . Furthermore , the use of poor man's continuation , where the last computed solution is used as an initial guess in the root-finding algorithm , means that we expect to see the same value of over a range of neighbouring values of the signal in this regime where we are looking at the first gradual loss of stability of the metastable basal solution . Once again , no meaning should be attributed to the clustering of values of in this case . In contrast to the other cases , the constitutively expressed fast HK transcription case only supports a basal solution for short time horizons: it is lost between and ( Fig . 4c ) . This broadly supports the arguments in the previous subsection , where the basal solution was found to be absent in the constitutively expressed fast transcription case and to persist for approximately in the remaining cases . The fact that the basal solution persists at all in the first case results from the stochastic nature of the simulations: there will always be a short delay in the formation of reporter protein when necessary chemical species are initially absent . At , we have and . Since at least one molecule of mRNA-HK , , is needed to initiate the reaction sequence that leads to the production of phosphorylated RR dimer , , and hence an induced level of reporter protein , , we expect the basal solution to persist for a time that is somewhat longer than , which in the constitutively expressed fast transcription case is approximately 3 s . The fact that the solution should persist for a somewhat longer time than 3 s results from stochastic delays in the formation of the intermediates ( phosphorylated HK dimer ) and ( RR-HK2P complex ) , which are also intially absent . This agrees reasonably well with the observed loss of the basal solution between and . The loss of the basal solution at a high external signal at a time horizon of only 500 s in the remaining cases is a little surprising , but we postulate that the solution corresponding to induced expression of the reporter gene is strongly attracting at high external signals and so small fluctuations might be enough to move a sufficient number of individual trajectories into the basin of attraction of this higher solution branch so that a mean basal solution would no longer exist . Once an average steady state is computed via Broyden's iterations , it is possible to calculate the corresponding Jacobian of the coarse time-stepper and infer the stability of the solution . Since the number of realisations used for the root-finding algorithm is relatively small ( ) the resulting Jacobian computations are affected by noise . At selected points on the bifurcation curve , we increased the number of realisations to and repeated the Jacobian computations times . In Fig . 5 we plot the spectra of the Jacobian evaluated at basal solutions for various values of the external signal . One instance ( out of the 10 calculations ) of each spectrum is plotted , except for the lower panel of Fig . 5c , where two instances are shown . In all four cases and for each of the 10 Jacobian computations , we found that solutions with low values of the external signal are stable . Conversely , high external signals lead to unstable steady states in the autoregulated fast and slow transcription cases and in the constitutively expressed slow transcription case . In the constitutively expressed fast transcription case we find a mixed picture for the high external signal: we repeated the Jacobian computation 20 times in this case and of those 11 gave a stable spectrum and 9 gave an unstable spectrum: one example of each is shown in the lower panel of Fig . 5c . We suggest that the difference in behaviour of the constitutively expressed fast transcription case compared to the other three may be due to the fact that the time horizon is much shorter - 50 s compared to 500 s - which could make the Jacobian calculations noisier , and that the steady state , which is effectively no longer a persistent basal solution , is further away from zero . Since the basal solution is expected to be only metastable at all values of the signal , we might have expected to see instability at low signals as well as high ones . However , in that region the higher solution branch - a true stable solution - lies close to the basal solution and so a ) it may be hard to separate the two within our given error tolerance and b ) the unstable eigenvalue of the basal solution will lie very close to the stability boundary and so we might classify it as stable within our given error tolerance . Broadly speaking a basal solution that appears stable at low external signals , but becomes unstable as the signal increases in strength , confirms our hypothesis of metastability . We can also investigate the existence of an induced expression solution using equation-free techniques . Here we start with a large external signal , and use a point in the vicinity of the solution predicted by the deterministic reaction rate equations ( 1 ) – ( 13 ) for induced expression of the reporter gene as an initial guess for a steady state . Once more we use poor man's continuation to follow the dependence of on the external signal , but this time tracking the solution as the external signal decreases . Any that we pick is likely to persist over sufficiently short time horizons , because we can pick a time interval so short that no reaction events are expected to take place . What we are really interested in are solutions that persist over long time horizons . However , once becomes greater than about 200 s , calculation times become so long as to be impractical . We would expect to find that for long enough , the autoregulated cases show ‘all-or-none’ behaviour where the activated expression solution suddenly vanishes below a threshold value of the external signal . In the language of nonlinear dynamics , this is a classical scenario of a subcritical bifurcation with hysteresis . By contrast for the constitutively expressed cases , we expect a smooth , graded , response as the external signal varies: in other words , a stable solution that grows in amplitude as the signal increases , but does not undergo a bifurcation . In the autoregulated cases , we do see an ‘all-or-none’ profile at the longest time horizon that we used , ( Figs . 6a and b ) . However , we actually see similar behaviour in the constitutively expressed cases ( Figs . 6c and d ) , though for the fast HK transcription case there is a hint of a graded response as the external signal decreases towards the point at which the basal solution appears . It is possible that the algorithm fails to converge on the induced steady state at intermediate values of the external signal , and instead locates the approximate basal solution . ( Even in the constitutive fast transcription case , this solution may occasionally be found to persist for 200 s owing to the stochastic nature of the system , and since the root-finding algorithm is permitted quite a large number of iterations it may pick it up . ) This may be because a larger ensemble of realisations is needed to achieve a given accuracy of solution as increases , as we describe below , but in practice using very large ensembles would have required infeasibly long run times . Interestingly we did find a graded response at in the constitutively expressed fast transcription case , but we lost it for lower values of the external signal when we increased the time horizon to ( see Fig . 7 ) . Perhaps this is indeed owing to the decreased accuracy in locating the solution . However , we note that another run at produced an ‘all-or-none’ profile ( not shown ) and that the autoregulated fast transcription case behaved similarly despite a graded solution not being expected there . Larger ensembles and longer run times would be necessary to resolve the question definitively . We have also computed the spectra corresponding to induced expression states for high values of the external signal , and found that they are stable in all cases ( see insets in Fig . 6 ) . In order to calculate the steady states , we repeatedly generate ensembles of realisations , each of which gives us a mean value . For a given , the variance of over a set of ensembles will be greater for longer and smaller ensembles . Thus , as increases , we really should use a larger ensemble of realisations to allow us to determine steady states with sufficient accuracy . It is likely that this would allow us to distinguish better between the behaviour in the constitutively expressed and autoregulated cases , but in practice this is computationally prohibitive . Furthermore , as we approach a steady state , the time evolution of a given trajectory becomes very slow ( because there is at least one growth-rate eigenvalue close to zero ) and so extremely long time horizons would be needed to identify the location of the steady state accurately . Nonetheless we do pick up the basal expression state at low values of the external signal and the induced expression state at high signals in all four cases , thus demonstrating the ability of the equation-free method to locate metastable and stable solutions in complex reaction networks where explicit analysis cannot be used , but where the time evolution of the system is accessible through a numerical time-stepper .
We have sought to explain the existence of the mixed-mode response in a stochastic kinetic model of the PhoBR TCS in E . coli . We used bifurcation analysis to show that this mixed mode was absent in the framework of deterministic reaction rate equations that govern the concentrations of chemical species in the thermodynamic limit , and that it must therefore result from stochasticity in the discrete system . We then analysed the discrete stochastic system directly using equations for the probabilistic mean number of molecules of each chemical species present , and showed that slow transcription of either or both of the histidine kinase or response regulator genes can lead the reporter gene to be expressed at basal level in a fraction of cells within a population , even when the external signal is so high that this would not be expected on deterministic grounds . We confirmed this finding using equation-free techniques that located the unexpected persistent basal expression state and ascertained that it is unstable at high external signals . This persistence of the basal level of reporter gene expression is a truly stochastic phenomenon that arises because we must wait until random processes lead both RR protein and phosphorylated HK dimer to be present in the cell simultaneously so that the chain of reactions that lead to the production of reporter protein can proceed . The delay will be lengthy if either transcription process is very slow , and that is why a basal level of expression can be observed over long times . Combined with a graded response to the signal in the case where the RR gene is constitutively expressed , the persistent basal state leads to the ‘mixed-mode’ response described by KZW [15] . When the RR gene is autoregulated , the persistent basal state effectively extends the range of external signals over which an ‘all-or-none’ response can be seen . These findings are important for understanding the survival strategies of bacterial pathogens . Two-component systems are the most prevalent mechanism of transmembrane signal transduction controlling gene expression programmes in bacteria [33] . Many of them are global regulators responsible for major switches in cell physiology . Thus stochasticity in the outcome of TCS regulation , that we have analysed in detail in this work , is likely to result in the coexistence of cells in qualitatively different physiological states . These cellular populations would inevitably respond differently to antibiotic treatment or immune system challenge and in many cases one of the populations would survive . Any global gene expression programme change leading to slow growth would slow down drug uptake and minimise the effects of drugs that block protein synthesis , and a change in the repertoire of surface proteins could enable a fraction of bacteria to survive an immune system attack . For example , a recent study shows involvement of the DosR response regulator in regulation of global metabolism and antibiotic response in M . tuberculosis . [34] Stochastic fluctuations in this particular TCS could therefore lead to the emergence of populations surviving antibiotic treatment . While the role of TCS stochasticity in pathogen survival has already been recognised [3] , the analysis of possible sources of phenotypic variation has been limited to autoregulated , bistable systems [3] . KZW did show numerical simulation trajectories exhibiting population diversity [15] , but they did not analyse the mechanisms underpinning the observed phenomena in detail . In this work we demonstrate for the first time that a TCS that is not multistable can generate bacterial population diversity at the timescales relevant to bacterial responses . The parameter configuration for which this behaviour is observed in our model is biologically plausible . Among the number of two-component systems studied in detail , both autoregulated ( e . g . PhoPQ in E . coli [35] ) and constitutive cases ( e . g . ArcAB in E . coli [36] ) have been observed . According to quantitative measurements [37] , the number of histidine kinase proteins present is low and could therefore be a source of stochastic fluctuations , as demonstrated in our model . In numerous two-component systems , such as ArcAB in E . coli [36] , the HK gene is expressed from a different transcription unit than the RR gene and in these cases low expression levels can be regulated at both the transcription and translation levels . Therefore , observed TCS architectures and measured protein amount ranges show that two-component systems exhibiting population diversity in the absence of autoregulation and multistability are likely to exist . Moreover , the observed diversity of TCS architectures shows that the two-component system is a highly evolvable regulatory network motif . Depending on point mutations in the promoter and ribosome binding site ( RBS ) , different modes of response to the external signal can be generated resulting in different distributions of phenotypic diversity in cellular populations . These mechanisms are likely to be subject to natural selection , especially in bacterial pathogens where population diversity conveys significant advantage . Our work shows for the first time that it is not only bistable two-component systems that are potential sources of phenotypic diversity in the evolution of bacterial populations . Thus experimental work on the stochasticity of two-component systems should not focus exclusively on multistable , autoregulated systems as has been the case so far . The analysis we have presented indicates that one should also consider the case where RR and HK genes are not autoregulated through positive feedback and where transcription of the HK gene is not coupled to the RR gene in an operon structure . Our study predicts that mutations in the promoter and RBS of this HK gene could result in population diversity and that the population would respond to the external signal in a mixed , rather than all-or-none fashion . Our work has also general implications for the understanding of molecular interaction networks other than two-component systems . We have analysed a large-scale model of a complete sequence of events linking external signal sensing with gene expression and shown the emergence of population diversity that does not derive from multistability of the system , but rather from slow production of a constituent chemical species . This phenomenon is very likely to be present in molecular interaction networks in general . The case of TCS histidine kinase indicates that noise in the expression of a single gene producing an external signal sensor can result in population diversity and a mixed-mode population response to that external signal . Potential occurrence of this mechanism should be taken into account in studies of a wide range of signal transduction cascades both in bacterial and eukaryotic cells . We show for the first time , to the best of our knowledge , the use of equation-free techniques to analyse a detailed model of a signal transduction and gene regulatory network . Our results demonstrate that this approach enables the application of the classical concepts of dynamical systems theory to the analysis of realistic stochastic models of molecular interaction networks of the cell . The calculation of the Jacobian is particularly useful as it provides insight into the stability of the behaviours observed in numerical realisations of stochastic dynamics . Understanding parameter dependencies in stochastic systems that are accessible only through direct numerical simulation is a major challenge . Hitherto , this has typically been attempted through time-consuming numerical experiments , without a systematic method for evaluating changes in the expected ( probabilistic mean ) system behaviour . Frequently , observation of a particular phenomenon in simulation trajectories brings little understanding of the underlying mechanism . Our work shows that equation-free methods provide a systematic and feasible solution to this problem . Our use of equation-free techniques to investigate stochastic phenomena in a biochemical reaction network of realistic scale demonstrates their potential for enabling greater insight into the behaviour of highly stochastic systems in biology , and also the challenges of scale that must be overcome in order to do so . To summarise , our work provides insight into the mechanisms of emergence of phenotypic diversity in populations of genetically identical cells . Our successful use of equation-free methods in this context will motivate future applications of this approach for the analysis of the stochastic dynamics of molecular interaction networks .
The rate of change with time of the vector of mean species numbers , , can be calculated from the chemical master equation ( 71 ) where is the probability that - see [24] , for example - and is the number of different types of chemical reaction in the system . The mean is given by , where is the number of chemical species , and so it evolves according to ( 72 ) where it should be noted that and are defined to be zero if for any . Note that if for some , then we have ( 73 ) and so since , and , , , we must have for all such that . Then for any function the mean is given by ( 74 ) The coarse time-stepper used in the equation-free method is composed from microscopic runs of the Gillespie algorithm in three stages:
|
It is a surprising fact that genetically identical bacteria , living in identical conditions , can develop in completely different ways: for example , one subpopulation might grow very fast and another very slowly . These different phenotypes are thought to be one reason why bacteria that cause disease can survive antibiotic treatment or become persistent . This diversity of behaviour is usually attributed to the existence of multiple stable phenotypic states , or to the coexistence of one stable state with another unstable excited state , or finally to the possibility of the whole biochemical system that controls the phenotype being switched on and off . In this paper we describe a different scenario that leads to phenotypic diversity in two-component system signalling , a very common mechanism that bacteria use to sense external signals and control their response to changes in their environment . We use probability theory and equation-free computational analysis to calculate the average number of molecules of each chemical species present in the two-component system and hence show that sporadic production of either of two key chemical components required for signalling can delay the response to the external signal in some bacterial cells and so lead to the emergence of two distinct cell populations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"systems",
"biology",
"mathematics",
"biology",
"computational",
"biology",
"signaling",
"networks",
"nonlinear",
"dynamics"
] |
2012
|
Equation-Free Analysis of Two-Component System Signalling Model Reveals the Emergence of Co-Existing Phenotypes in the Absence of Multistationarity
|
Introgressing anti-pathogen constructs into wild vector populations could reduce disease transmission . It is generally assumed that such introgression would require linking an anti-pathogen gene with a selfish genetic element or similar technologies . Yet none of the proposed transgenic anti-pathogen gene-drive mechanisms are likely to be implemented as public health measures in the near future . Thus , much attention now focuses instead on transgenic strategies aimed at mosquito population suppression , an approach generally perceived to be practical . By contrast , aiming to replace vector competent mosquito populations with vector incompetent populations by releasing mosquitoes carrying a single anti-pathogen gene without a gene-drive mechanism is widely considered impractical . Here we use Skeeter Buster , a previously published stochastic , spatially explicit model of Aedes aegypti to investigate whether a number of approaches for releasing mosquitoes with only an anti-pathogen construct would be efficient and effective in the tropical city of Iquitos , Peru . To assess the performance of such releases using realistic release numbers , we compare the transient and long-term effects of this strategy with two other genetic control strategies that have been developed in Ae . aegypti: release of a strain with female-specific lethality , and a strain with both female-specific lethality and an anti-pathogen gene . We find that releasing mosquitoes carrying only an anti-pathogen construct can substantially decrease vector competence of a natural population , even at release ratios well below that required for the two currently feasible alternatives that rely on population reduction . Finally , although current genetic control strategies based on population reduction are compromised by immigration of wild-type mosquitoes , releasing mosquitoes carrying only an anti-pathogen gene is considerably more robust to such immigration . Contrary to the widely held view that transgenic control programs aimed at population replacement require linking an anti-pathogen gene to selfish genetic elements , we find releasing mosquitoes in numbers much smaller than those considered necessary for transgenic population reduction can result in comparatively rapid and robust population replacement . In light of this non-intuitive result , directing efforts to improve rearing capacity and logistical support for implementing releases , and reducing the fitness costs of existing recombinant technologies , may provide a viable , alternative route to introgressing anti-pathogen transgenes under field conditions .
The mosquito-borne dengue virus , transmitted primarily by the yellow fever mosquito Aedes aegypti ( Linnaeus ) is estimated to cause approximately 390 million infections each year ( [1] ) . No widespread prophylactic treatments are currently available for dengue , and thus vector population control remains the primary public health strategy . The release of genetically modified mosquitoes provides one approach towards vector population control . Releasing mosquitoes that carry transgenes rendering them vector-incompetent could , in principle , facilitate the prevention of epidemics by replacing wild-type populations of Ae . aegypti with mosquitoes carrying transgenic anti-pathogen constructs ( e . g . , [2] ) . Ultimately , the frequency of such anti-pathogen transgenes must increase to fixation ( e . g . , [3] ) , or at least reach and remain at a sufficiently high frequency to lower the number of competent vectors to levels preventing epidemic outbreaks ( e . g . , [4] , [5] ) . Aedes aegypti populations in cities are potentially quite large , and the large release numbers generally perceived to be necessary for population replacement has inspired the development of several transgenic gene-drive strategies and related approaches . These approaches including linking anti-pathogen constructs to MEDEA ( [6] ) or homing endonuclease genes ( [7] and [8] ) , inducing widespread Wolbachia infections ( e . g . , [9] , [10] and [11] ) , among others ( reviewed in [3] and [12] ) . These gene-drive strategies vary in their stages of development and future potential for application in public health contexts . Only those based on Wolbachia symbionts have thus far been demonstrated to be workable in the field for mosquitoes ( [13] ) . Even for Wolbachia-based interventions , the epidemiological impact of this approach under realistic conditions remains unknown . For instance , the epidemiological performance of Wolbachia-induced refractoriness , particularly in regions where multiple serotypes potentially circulate ( e . g . , [14] and [15] ) remains to be tested . Introducing anti-pathogen genes into wild populations therefore remains a major challenge . Moreover , because any specific anti-pathogen gene for reducing disease transmission in the field could fail due to pathogen resistance evolution , a system that enables effective replacement of one anti-pathogen transgene with another is critical to the long-term success of population replacement programs . As none of the proposed transgenic anti-pathogen gene-drive mechanisms are likely to be implemented as public health measures in the near future ( [16] , [12] and [17] ) , transgenic strategies aimed at suppressing mosquito populations have received considerable attention ( [18] , [19] , [20] , and [21] ) . These approaches are inspired by the sterile-insect technique ( SIT ) , whereby a large number of sterile males are reared , irradiated and released , subsequently mating with wild females and thereby reducing the fecundity of wild female mosquitoes ( [22] , [20] ) . Although traditional SIT programs for Ae . aegypti have never been implemented over large geographic regions or for extended periods of time ( [23] and [24] ) , transgenic population reduction strategies in Ae . aegypti have advanced to field trials ( [25] and [26] ) . Our examination of the literature indicates that there is a prevailing view that population suppression by repeated release of transgenic , sterile mosquitoes may be practical ( [18] , [24] ) . By contrast , aiming at population replacement by repeated release of mosquitoes carrying a single anti-pathogen gene without a gene-drive mechanism is generally not seen as practical ( e . g . [3] , [27] and [28] ) , unless the transgenic construct can provide a net fitness benefit to refractory mosquitoes ( e . g . , when the virulence of the pathogen to mosquitoes is high - [29] and [30] ) . Yet despite this prevailing view , there are surprisingly few quantitative assessments of how releasing anti-pathogen genes into a mosquito population without an accompanying gene-drive mechanism could alter the genetic composition of natural Ae . aegypti populations . Quantifying the effectiveness of release strategies aimed at population replacement that do not rely on gene-drive mechanisms under realistic conditions is key to establishing a baseline against which the efficacies of more elaborate drive mechanisms can be compared ( [29] ) . Such an assessment can potentially help evaluate the performance of existing anti-pathogen genes under field conditions ( e . g . , [27] ) , as well as identify key facets of mosquito biology that might promote or hinder the spread of anti-pathogen constructs in the absence of gene-drive ( e . g . , mating preferences among transgenic and wild-type mosquitoes - [30] ) . Here we use Skeeter Buster ( [31] ) , a biologically detailed model of Ae . aegypti population dynamics parameterized for a tropical city , to quantitatively assess whether a transgenic control program based on rearing and releasing transgenic Ae . aegypti carrying a single anti-dengue gene ( e . g . , [2] ) could provide a feasible approach to population replacement in the absence of gene-drive mechanisms . We compare our results to release sizes necessary for suppressing vectoring mosquito numbers with a transgenic strain carrying 1 ) a single anti-pathogen gene , 2 ) a single female-killing gene , and 3 ) both an anti-pathogen gene and a female-killing gene ( e . g . , [32] ) .
Skeeter Buster models the population genetics and dynamics of Ae . aegypti in an urban setting , incorporating key processes in the life cycle , including temperature-dependent survival rates , container-level development and nutrient dynamics , oviposition , and dispersal ( [31] ) . Four life stages are explicitly modeled: eggs , larvae , pupae , and adults . Development , reproduction and mortality are stochastic , but these processes are also driven by temperature- and , for development , resource-dependent rates . Skeeter Buster models individual water-holding containers located in specific houses ( “sites” ) laid out on a rectangular grid . Resource dynamics within containers and the feedback between larval biomass and resource availability ( i . e . , density dependence ) in containers with mosquitoes in them are based on the equations used in [33] . Mosquitoes emerge as adults and occupy sites where the containers in which they developed are located . Females select mates among males in the same site , mate only once during their lives , and oviposit in containers at the site they occupy on a given day during each gonotrophic cycle . Adults can potentially migrate each day to a randomly selected adjacent site ( fixed daily probability of migration = 0 . 3 based on the mark-release recapture studies of [34] and parameterized according to [31] , [35] ) . In Skeeter Buster , females may also undergo occasional long range dispersal events ( for instance , via inadvertent translocation in vehicles ) . Such long distance dispersal is modeled by allowing a small proportion ( 2% ) of individuals to disperse to a site randomly selected within a Manhattan distance of twenty sites . The resulting dispersal pattern has been shown to be consistent with field studies ( e . g . , [36] and [31] ) . For a more thorough description and justification of the features and components of the model , see [31] , [35] , and [37] . Because Skeeter Buster aims to model the biology of Ae . aegypti in a realistic environment , it must be parameterized in a location-specific manner with local meteorological data and a description of the distribution of containers . Data from the equatorial Amazonian city of Iquitos , Peru ( 73 . 2W , 3 . 7S ) , which has been subject to long-term , detailed larval habitat surveys ( e . g . , [38] ) , has proven particularly well-suited for use with Skeeter Buster ( e . g . , [31] , [35] and [37] ) . Although there is currently no planned release of transgenic mosquitoes in Peru ( A . Morrison , pers . comm . ) , using Iquitos as a case-study allows us to build on this earlier work and apply Skeeter Buster to model the population dynamics of Ae . aegypti at this location ( e . g . , [37] ) . As in [39] , a 2448 house subsection ( 68 houses ×36 houses ) of the city is simulated with periodic boundary conditions , following the approach described in [37] . In Iquitos , the average distance between sites is on the order of 5-10 meters ( [37] ) , and the simulated region corresponds to an area of approximately one square kilometer of a densely populated neighborhood in Iquitos . These conditions result in an equilibrium population size of approximately 14000 total female adult mosquitoes ( or approximately 6 female adult mosquitoes per site ) prior to releases of transgenic mosquitoes . Skeeter Buster follows the genotypes of mosquitoes , allowing evaluation of changes in the genetic composition of the population that result from releasing transgenic mosquitoes . This feature permits quantitative comparisons between our results and those obtained for alternative genetic control strategies . The anti-pathogen and female-specific lethal genes are both modeled as diallelic loci . Each locus is characterized by the presence or absence of the transgenic construct . For brevity , we refer to individuals that lack both constructs as “wild-type mosquitoes” . We simulate introgression by the anti-pathogen transgene when mosquitoes carrying only the anti-pathogen gene are released ( for brevity , we denote this as the “AP” strategy ) . We compare the resulting reduction in competent vectors to reductions caused by a transgenic control program based on the female-specific lethal gene alone ( an “FK” strategy aiming exclusively at population reduction ) , as well as to a “reduce and replace” ( “RR” ) strategy based on releasing mosquitoes carrying both the anti-pathogen and female-specific lethal transgenes . As an FK strategy permits the introduced female-specific lethal transgene to remain in the population past the first generation , it can potentially accelerate population reduction relative to non-sex specific lethal constructs ( e . g . , [40] ) . To highlight the contrast between transgenic strategies based on population reduction and approaches aiming at population replacement , we do not analyze population reduction strategies that cause both male and female offspring mortality ( for instance , bisex RIDL and conventional SIT strategies - e . g . , [20] and [25] ) . The female-specific lethal gene is assumed to be expressed only when eggs are reared in the absence of tetracycline ( [19] ) . Inserting transgenes into the Ae . aegypti genome may potentially disrupt the normal functioning and regulation of the inserted site . We assume such fitness costs may reduce the expected survivorship of transgenic individuals relative to wild-type individuals by 5-10% per copy of the anti-pathogen transgene carried ( the midpoint and upper limit of the range used in , e . g . , [27] and [41] ) . For simplicity , and to facilitate our comparison of the different strategies , we do not model any fitness costs associated with the female-specific lethal gene aside from female-specific conditional mortality . We also assume that when present , costs associated with the anti-pathogen gene are expressed at the egg stage for both male and female mosquitoes , and hence potentially result in a reduction in larval production when either parent carries a transgene copy ( [42] ) . We compare how the presence and absence of such fitness costs associated with the transgene insertion affect prospects for population replacement . While the dengue virus may confer a fitness cost on infected mosquitoes ( e . g . , [43] ) , how much evading infection facilitates the transgene's spread depends , in part , on dengue prevalence among vectors ( [29] ) . This prevalence , in turn , can be highly location- and time-specific , and governed by several factors including local epidemiological dynamics , the susceptibility of mosquitoes to vertical transmission of dengue , as well as potentially seasonal effects ( e . g . , [44] , [45] and [46] ) . Incorporating all these processes into our model is beyond the scope of our study . Thus , for simplicity and to stringently test the effectiveness of population replacement strategies absent gene-drive , we assume throughout that mosquitoes carrying the anti-pathogen gene enjoy no fitness benefit relative to wild-type mosquitoes . Finally , genetic control programs relying on population reduction alone are vulnerable to reinvasion if releases cease . Wild Ae . aegypti individuals from populations that have not been subject to control efforts can immigrate into the population , undermining sustainable population reduction in the absence of ongoing releases ( [47] and [48] ) . By contrast , the effects of immigration on the efficacy of population replacement strategies are relatively understudied . [39] shows how an RR strategy can also be subject to such adverse effects of immigration . Comparing the robustness of the different strategies in the face of immigration by wild-type mosquitoes can therefore be critical in evaluating their potential under biologically relevant conditions . We thus assess the impact of gravid wild-type immigrant females on the efficacy of the population replacement programs we consider . All released transgenic individuals are modeled as homozygotes carrying two copies of the anti-pathogen transgene . The anti-pathogen gene is considered to have dominant phenotypic expression . Because transgenic female mosquitoes carrying an anti-pathogen gene do not transmit dengue , we compare male-only adult releases ( which may be more readily amenable to regulatory and community approval e . g . , [49] and [50] ) to bi-sex adult releases that would increase nuisance biting to some extent . Releases of genetically-modified mosquitoes are modeled by dynamically adding cohorts of mosquitoes homozygous for the anti-pathogen gene at specified dates to specific sites following a single burn-in year permitting the population to attain demographic equilibrium ( e . g . , [51] and [39] ) . Distributing and releasing transgenic adults can be logistically challenging . For instance , the adult-stage of Ae . aegypti is comparatively short-lived in the field , and thus releasing adult mosquitoes may require planning releases to coincide with the emergence of individuals as adults . An alternative strategy is to release transgenic mosquito eggs . Ae . aegypti eggs are desiccation-resistant and can survive for extended periods of time ( e . g . , [52] , [53] ) , permitting long-term storage and subsequent distribution . Using viable eggs has therefore been suggested as an alternative to releasing adults ( [54] , [24] and [51] ) . Eggs could also potentially be easier to handle , reducing the costs associated with transporting transgenic strains . We therefore evaluate the effectiveness of a release program involving the distribution of transgenic eggs instead of adults . Following [19] , [51] and [39] , the release of eggs is modeled as weekly additions into specified sites of nutrient-filled containers shielded from ovipositing females . We model 100 to 1600 eggs placed in covered , nutrient-filled containers . Although sufficient resources are provided in the containers , only approximately 40% of the eggs ultimately develop to adulthood due to natural mortality . These added containers are removed from the simulation upon the emergence as adults or death of all mosquitoes they contain . Because male and female eggs cannot currently be separated , an equal number of eggs of both sexes are assumed to be placed in each of the containers . We base the release ratios for mosquitoes on those used in an earlier study ( [51] ) employing Skeeter Buster to evaluate how releasing transgenic mosquitoes carrying a single conditionally-lethal construct could facilitate population reduction . The largest release numbers for which we present results ( totaling approximately 120 , 000 adult males per week ) corresponds to a release ratio of approximately 17 . 5∶1 released males to wild-type males present when releases begin . These ratios are within the range used in control programs for mosquitoes based on the sterile insect technique ( e . g . , [55] ) . [51] found that the weekly release numbers necessary to cause local population extinction depended on the geographic penetrance of the release regime . Furthermore , the ability to implement the transgenic control programs characterized above could also depend on the accessibility of different sites in the urban arena . Release sites may be clumped in space in some instances , whereas in other cases it may be feasible to release individuals at regularly-spaced points across the grid , or , through aerial releases of adults , potentially into a very high proportion of sites . To approximate the potential effect of different spatial implementations , we model three idealized spatial configurations: homogenous releases everywhere , regular point-source releases at every site at regularly-spaced points in the grid , and random point-source releases at 10% of the sites randomly chosen . None of these regimes can be expected to be achieved exactly under field conditions , but comparing the distinct spatial patterns can help inform the spatial configuration of releases that actual transgenic control programs should strive to reproduce . For the operational reasons explained in [51] , we do not examine the distribution of eggs into every site . In the cases where releases only occur in a subset of sites , we fix the sites at which releases take place for the duration of each simulation run . If the composition of release sites changes frequently as the releases are carried out , then the release regime can be expected to begin to approximate the homogenous configuration over time . Such homogenization can make it difficult to contrast the effects of releasing transgenic mosquitoes at a limited number of sites to the case where releases occur at all sites . Following [51] and [39] , we assume indefinite releases cannot be sustained . Thus , we compare a relatively short ( one year ) release period to a longer ( three year ) release period . To assess the fate of the transgene after releases end , we run the model for an additional two years . Finally , because Skeeter Buster is a stochastic model , we run 30 simulations of each release regime varying the spatial arrangement of sites ( and hence containers ) in the grid for each simulation . For the random point-source releases , we also randomize the release sites at the beginning of each simulation .
Across all release scenarios , releases of transgenic mosquitoes under a “replacement-alone” ( AP ) strategy ( mosquitoes carrying an anti-pathogen gene without a conditionally-lethal construct ) lowered the long-term population of vector-competent mosquitoes more successfully than the “reduce and replace” ( RR - mosquitoes carrying both the anti-pathogen and female-lethal construct ) strategy ( Fig . 4 ) . In comparison to the RR releases , even in the presence of a fitness cost , releases of mosquitoes under the AP strategy proved capable of maintaining the vector-competent population at low levels . We note that these simulations assume the female-specific lethal construct provides no additional fitness cost ( e . g . , male carriers of the lethal gene are unaffected ) . When the female-killing transgene is likely to carry to a fitness cost , the reduction in competent vectors for the RR ( and FK ) strategy may be considerably less than what we show here ( e . g . , [51] ) . The contrast between the two transgenic control strategies can be particularly pronounced when releases last for three years and the anti-pathogen gene carries a fitness cost . For an RR strategy , the detrimental effects of a fitness cost can become more apparent especially during population recovery as the entire mosquito population grows rapidly from low numbers . This results in the competent vector population size quickly recovering towards pre-release levels . By contrast , even in the presence of a fitness cost , releases of mosquitoes carrying only the anti-pathogen gene result in low numbers of vector-competent mosquitoes for extended periods of time following the end of releases , and the vector competent population recovers much more slowly . We also find transgenic control strategies based on population reduction alone ( the FK strategy ) fail to lower the long-term vector competent population sizes below the levels obtained using the RR and AP strategies , even when the anti-pathogen construct carried a fitness cost ( Figure S2 ) . The AP strategy consistently lowered the long-term number of vector-competent mosquitoes more than the RR strategy . Under an RR strategy , population reduction can release the surviving mosquitoes from density-dependent constraints , allowing surviving wild-type mosquitoes to have high per-capita growth rates . By contrast , the AP strategy does not lower population density , and hence density dependence can remain strong even as releases are ongoing . Under some conditions ( e . g . , when only a small number of adult males are released Fig . 4A ) , this difference between the two strategies prevents the RR strategy from being able to lower the number of vector competent mosquitoes as effectively as the AP strategy once the vector competent population has been reduced , even when releases are ongoing ( e . g . , Figs . 4A , 4C , and 4D ) . Thus , for some release scenarios ( e . g . , releases of eggs ) , an AP strategy can have a larger transient effect on vector-competent population reduction than an RR strategy ( e . g . , Fig . 4C and Fig . 4D ) . By contrast , when the RR strategy is comparatively effective at reducing mosquito abundances ( e . g . , when sufficiently large numbers of females are also released - e . g . , Fig . 4B ) , an RR strategy can cause greater reductions in the vector-competent mosquito population during the transient stages in some of the release scenarios ( Figs . 4 and 5 ) . We find that , when present , these differences are most pronounced after there has been an approximately three orders of magnitude decline in the vector-competent population caused by an RR strategy , although the difference can be apparent by 200 days into releases ( e . g . , Fig 5A ) . When the fitness cost is very high , the difference between an RR strategy and an AP strategy during the transient stages has the potential to be modestly larger than when there is weaker or no fitness cost , because the higher fitness cost renders an anti-pathogen construct less capable of spreading through an AP strategy . An FK strategy based on population reduction alone proves unable to lower the number of vector competent mosquitoes further than the RR strategy , even during the transient stages ( Figure S2 ) . These results are robust to whether simulation runs resulting in population extinction are included in the analysis ( Figures S3-S4 ) . We find that immigration by gravid , mature wild-type females has an appreciable effect on the long-term numbers of vector competent females under an AP transgenic control strategy . As immigration from wild-type populations increases ( e . g . , at least 5 immigrants per day across the simulated region ) , the frequency of wild-type mosquitoes can also increase several fold ( Fig . 6A ) . Increasing the number of AP-only mosquitoes released has only a small effect on these results , as does eliminating the fitness cost associated with the anti-pathogen transgene . As the immigration rate increases , the effect of increasing release numbers on the number of vector competent mosquitoes becomes more apparent ( black versus grey lines , Fig . 6A ) . Although immigration can potentially increase the number of competent vectors , its impact on transgenic control strategies other than AP can be more pronounced . For instance , under an RR strategy , even comparatively rare immigration events severely undermine population replacement efforts ( Fig 6B; see also [39] ) . Moreover , when the transgenic control strategy involves population reduction , increasing the numbers released can actually accentuate the effect of wild-type immigration events ( Fig 6B; see also [39] ) .
Because the genetic engineering of anti-pathogen constructs for Ae . aegypti is somewhat further developed than for gene-drive mechanisms ( e . g . , [27] , [12] and [17] ) , it is critical to assess the prospects of alternative approaches to spreading currently proposed anti-pathogen constructs under field conditions . Such an assessment could provide a baseline against which the field efficacies of gene-drive mechanisms and related approaches could be compared ( e . g . , [29] ) . As genetic control methods based either in whole or in part on population reduction may require considerable resources to sustain , strategies aimed at population replacement provide an alternative approach to achieving long-term reductions in vector competence . Based on simulations with a stochastic , spatial model of a natural population of Ae . aegypti , we find that releasing mosquitoes carrying only a single anti-pathogen construct at ratios well below those considered necessary for transgenic technologies based on population reduction can facilitate robust reductions in vector-competence in a reasonable time frame , in some cases reducing the average number of competent vectors to between to of pre-control levels ( e . g . , Figs . 1 and 2 ) . These reductions compare favorably to reductions in vector capacity considered necessary to achieve public health goals . For instance , reducing vector capacity ( as measured via house indices ) using source removal , space spraying and legal and educational interventions to between to of pre-control levels in Cuba and Singapore facilitated dengue control in both countries in the 1980s ( [56] ) . However , we caution that the ultimate epidemiological benefits of any transgenic control program depends on the effectiveness of the anti-pathogen transgene under field conditions . When transgenic females are unable to carry dengue , we also find that releasing females in addition to males greatly reduces the number of mosquitoes necessary to reduce vector competence ( Figure 1 , Figure 3 and Figure S1 ) . These results appear robust across a range of release regimes . In particular , we also show that releasing even very few eggs per house ( especially in comparison to the number of eggs that may need to be distributed to cause population elimination - e . g . , [51] ) to be quite effective . A transgenic control program based on distributing eggs will not require timing release events to coincide with adult emergence events , and may be logistically easier to implement or prove more cost effective . Nevertheless , improving the geographic uniformity of releases for both adult and egg releases facilitates introgression , particularly when transgenic mosquitoes bear fitness costs . Finally , our comparison to other genetic control strategies shows that an AP strategy is considerably more robust to immigration than the RR strategy . Under the RR strategy , increasing release numbers results in a trade-off between population replacement and vulnerability to immigration; by contrast , an AP strategy implies no such trade-off , and the effects of wild-type immigration can be reduced by releasing more mosquitoes carrying only an anti-pathogen construct . A frequently cited limitation to successfully introgressing an anti-pathogen gene ( e . g . , [2] ) without an accompanying gene-drive mechanism is that such an approach may require prohibitively large release numbers ( e . g . , [57] , [3] , [27] , [58] , [12] ) . However , based on our results modeling the Ae . aegypti population in a neighborhood of approximately 2500 houses in Iquitos , Peru , total weekly release numbers of less than 25000 adult male and female mosquitoes into the expected population of approximately 21000 adults suffice to severely reduce the number of vector-competent females two years after releases end . The total release numbers we simulate are comparable to the number of mosquitoes used to establish Wolbachia in the trial studies in Yorkey's Knob and Gordonvale , Australia , communities of approximately 615 and 670 houses , respectively ( [13] ) . There , between 10 , 000–22 , 000 Wolbachia-infected mosquitoes were released weekly into a seasonal mosquito population ( [13] ) ( or roughly 50–110 mosquitoes per hectare per week , assuming a combined release area of approximately 200 hectares across the municipalities – Text S1 ) . In seasonal environments , such as the communities where the Wolbachia-infected mosquitoes were released , releasing mosquitoes as the population is increasing from a seasonal minimum could improve the efficacy of transgenic control strategies using Wolbachia ( [59] ) . Yet even for a relatively stable , non-seasonal mosquito population ( as in Iquitos ) , our results suggest that the release densities required for a successful AP strategy may be quite modest . Assuming an average household size of 5 . 8 people in Iquitos ( e . g . , [60] ) and a land area of approximately 78 , 400 hectares for the entire city ( [61] ) , our simulated region represents an area of Iquitos containing roughly 4% of the human population . A crude extrapolation of our results suggests that approximately 650 , 000 mosquitoes would need to be released weekly ( for a total of approximately 34 million individuals released per year ) throughout an entire city the size of Iquitos to render a substantial fraction of Ae . aegypti vector-incompetent at the end of the release duration . This translates into a release density of 8 to 9 released mosquitoes per hectare , per week . By comparison , a transgenic SIT-based release program that reduced the population size of Ae . aegypti at a site in the Grand Cayman Islands by about 80% involved approximately 3 . 3 million engineered males released in a 23-week period ( [25] ) , or roughly 143 , 000 male mosquitoes per week . This was the equivalent of about 3150 males per hectare , per week , and [25] note that these release rates constitute the minimum necessary to cause population elimination in the absence of immigration . This represents a figure several orders of magnitude larger than the release numbers that we found necessary for introgressing an anti-pathogen gene without an accompanying gene-drive mechanism in our study . Our results therefore suggest that transgenic control strategies based on population replacement could be plausibly implemented to reduce vector capacity even in the absence of mature gene-drive like technologies . Our comparative modeling approach allows us to highlight why releasing mosquitoes carrying only an anti-pathogen construct ( the AP strategy ) can lower vector competence more effectively than approaches based on population reduction ( the “reduction-only” FK strategy and a “reduce and replace” RR strategy ) . Skeeter Buster has previously been applied to evaluate transgenic control strategies that rely on population reduction ( [51] for an FK strategy , and [39] for an RR strategy ) . These studies found that some wild-type genes can be expected to persist due to inherent stochasticity in the simulation runs . [39] show that when mosquitoes that do not carry the anti-pathogen gene are able to persist following large population reductions , genetic drift at small population sizes can hinder the effectiveness of the RR strategy . The effects of genetic drift at small population sizes in the RR strategy can be reflected in the genetic composition following population recovery , rendering sustainable reductions in vector competence using an RR strategy very challenging ( [39] ) . By contrast , the AP strategy does not cause population reduction . Continually releasing mosquitoes carrying an AP transgene can therefore monotonically increase the transgene's frequency ( e . g . , Fig . 1 ) . Furthermore , the AP strategy does not cause population reduction , and thus wild-type mosquitoes remain subject to density-dependent pressures , reducing their ability to contribute heavily to the genetic composition of subsequent generations . These mechanisms allow the AP strategy to perform better than the RR strategy after releases end under a wide array of release scenarios , and improve the robustness of the AP strategy to immigration in comparison to the RR strategy . However , we note that , at least during the transient stages , the RR strategy can lower vector competence further in some instances compared to the AP strategy . In our simulations , we find such transient effects to be most apparent when the population of vector-competent mosquitoes has already been reduced substantially . Thus , the public health benefits of higher transient reductions could be limited , while the risk can be much more pronounced , particularly considering the long-term failure in the face of immigration of wild-type mosquitoes . Nevertheless , such enhanced ( albeit transient ) reductions caused by the RR strategy may justify switching between alternative transgenic strains as releases are ongoing . In a subsequent paper , we analyze whether switching between released strains could exploit transient reductions to maximize the potential for long-term reductions in vector competence ( Robert et al . , in review ) . Our work also provides a framework that allows communities and other stakeholders to assess the benefits and costs of implementing different release regimes and genetic control strategies . We anticipate individuals and entities to differ in their willingness to sustain a prolonged genetic control program of Ae . aegypti ( [62] ) . Reductions in vector-competence they consider necessary to reduce transmission of Ae . aegypti-vectored diseases may depend on , among other factors , receptiveness to genetically-based control strategies ( e . g . , [23] ) or existing public health strategies ( e . g . , source removal or clinical interventions ) that could complement a genetically-based vector control program . Our approach presents a quantitative basis for characterizing both the anticipated reduction in vector competence and the corresponding release numbers that would be required to attain such reductions . Potentially , the costs of implementing a transgenic release program could then be compared to the costs associated with implementing alternative disease management strategies . Cost-benefit analyses are particularly critical when releases of transgenic females are being considered . The viability of this approach requires carefully weighing real as well as perceived risks in the affected communities . Female releases may be unacceptable unless the anti-pathogen gene renders its carriers completely vector incompetent . Additionally , in communities at risk for other diseases vectored by Aedes aegypti , assessing how effectively the transgene protects against other pathogens ( e . g . , chikungunya - [63] ) will be critical . Such risks must be weighed against the cost savings of working with fewer transgenic mosquitoes . Evaluating the marginal effects of producing additional mosquitoes required for male-only releases is key to such an assessment . Our approach allows comparing the number of transgenic mosquitoes required under male-only and bi-sex releases to obtain a given reduction in vector competence . Communities could then weigh the cost of risk mitigation necessary for female releases ( e . g . , investing in improved anti-pathogen constructs ) against the marginal costs of producing and releasing more mosquitoes required for male-only releases . Our analyses sought to compare the efficacies of distinct transgenic release strategies in a tropical urban environment . However , we calibrated our model for the Ae . aegypti population in Iquitos , Peru , and location-specific assessments should precede implementation of a transgenic vector control program in other localities . Such analyses may suggest that some Ae . aegypti populations are less amenable to the transgenic strategies we consider . For example , the distribution of breeding containers in other tropical cities may be more spatially heterogeneous than the distribution in Iquitos ( [38] and [64] ) . Under the prevailing pattern of container distribution in Iquitos , there appears to be little spatial clustering in the predicted distribution of female mosquitoes carrying an anti-pathogen transgene ( Figure S5 ) . In other localities , if breeding containers are much less evenly distributed than they are in Iquitos , then the spatial distribution of the anti-pathogen construct may differ . Nevertheless , as [51] note , high levels of spatial heterogeneity can also be expected to reduce the efficacy of transgenic control programs based on population reduction . Thus , how increased spatial heterogeneity differentially affects transgenic strategies based on population reduction and strategies based on population replacement may therefore vary by location . Our framework provides a potential approach to compare such effects across different communities . Our results may also be applicable should it become necessary to eliminate the transgene ( see also [27] ) or to prevent its spread . Presumably wild-type male mosquitoes could be released at numbers comparable to those described here ( or possibly lower , especially if wild-type mosquitoes have a fitness advantage ) in order to eliminate the transgene . Alternatively , should pathogens evolve resistance to the vector's expressed anti-pathogen mechanisms or to become more virulent to humans ( e . g . , [65] ) , we anticipate that releases using alternative or complementary transgenic constructs ( [27] ) may be considered even without the need to successfully link all such constructs to a gene-drive mechanism . However , releases with multiple constructs may also raise the effective fitness costs experienced by transgenic strains substantially . Following previous modeling studies ( e . g . , [27] , [41] and [39] ) , we considered an additive fitness cost of 5–10% per copy of the anti-pathogen gene . We found that although such a fitness cost could reduce the long-term frequency of vector-incompetent females , the rate at which the vector-competent population recovered was relatively mild ( e . g . , Fig . 1 ) especially when compared to the rate of recovery of the vector-competent population in a RR or FK-only strategy ( e . g . , Fig . 4 ) . Nevertheless , some lab- and field cage-based studies have reported that fitness costs associated with transgenic insertions in Ae . aegypti can potentially be much larger or operate at different life stages ( e . g . , [66] and [67] ) . How such fitness costs are expressed under field conditions remains an open question ( e . g . , [68] ) , although improvements have been made in reducing fitness costs at least in artificial settings ( [69] ) . As specific anti-pathogen constructs become available , our comparative framework provides one approach to assessing the release numbers necessary to compensate for different reported fitness costs . Such comparative assessments could be used to discern if the release numbers required are prohibitive , and can provide quantitative evidence for the need to invest in developing gene-drive like technologies , further reducing the fitness cost , or alternative anti-pathogen transgenes . Finally , in some cases the dengue virus may confer a fitness cost on infected mosquitoes ( e . g . , [43] ) , providing a potential fitness benefit to transgenic mosquitoes that can facilitate spread without being linked to a gene-drive mechanism ( [29] ) . Our results indicate that even without conferring a fitness benefit , an anti-pathogen transgene may substantially reduce vector competence in an urban Ae . aegypti population . However , if carrying a transgene provides a net fitness benefit to mosquitoes in the field , then under certain conditions ( e . g . , if there is sufficient assortative mating among transgenic mosquitoes - e . g . , [30] ) , releasing more mosquitoes might not be necessary to compensate for fitness costs associated with the transgene . In light of our results that population replacement may not require gene-drive , assessing any potential fitness benefits transgenic constructs provide their bearers , as well as quantifying the refractoriness provided by the anti-pathogen construct , become particularly salient issues . Some transgenic control programs , particularly those based on population reduction , frequently apply a density-independent control method or wait for seasonally low densities before conducting releases ( [70] ) . For instance , [51] simulated a two-week , pre-release vector control program using a traditional method ( e . g . , insecticide spraying ) that lowered the number of released mosquitoes carrying a dominant , conditionally lethal-construct necessary to achieve extinction . Some preliminary analyses show , in principle , that a similar approach could potentially improve prospects for population replacement , although the duration of the traditional intervention needs to be somewhat longer to have a detectable effect when release sizes are small ( Figure S6 ) . A possible further advantage of having population reduction precede transgenic population replacement efforts ( rather than aim to simultaneously reduce and replace a vector population ) is that it could decouple an anti-pathogen construct from the conditionally-lethal gene . This could mitigate any detrimental effects transient linkage disequilibria may have when the two constructs are released simultaneously ( as may occur in the RR strategy - [32] ) . Using a pre-release control that can rapidly cause significant population reductions ( e . g . , spraying ) may therefore help raise the effective release ratio and provide an alternative to implementing a genetic replacement strategy over an extended period of time or , where rearing or fitness costs are potentially high , allow even fewer mosquitoes to be released than the numbers considered here . In conclusion , our results raise an intriguing possibility: even in the absence of drive mechanisms or fitness advantages conferred by transgenic constructs , releasing mosquitoes in numbers much smaller than those considered necessary for other genetic management strategies ( e . g . , those based on population reduction ) can result in rapid and robust population replacement . Although smaller release numbers may suffice to establish Wolbachia in some settings ( e . g . , [71] ) , alternative , engineered anti-pathogen genes may continue to be necessary in responding to possible failures of Wolbachia ( or any other specific anti-dengue construct ) to reduce disease transmission in the field . Directing efforts to improve rearing capacity and logistical support for implementing releases , and reducing the fitness costs of existing recombinant technologies , may provide a viable , alternative route to introgressing anti-pathogen transgenes under field conditions .
|
Dengue is transmitted by the Aedes aegypti mosquito . Releases of genetically sterile males have been shown to reduce wild mosquito numbers . An alternative approach is to release mosquitoes carrying genes blocking dengue transmission . It is often assumed that spreading such genes in mosquito populations requires using selfish genetic elements ( SGEs - genes that are inherited at higher rates than other genes in the genome ) . Absent such techniques , the release numbers required to transform mosquito populations is seen as prohibitive . However , strategies that rely on SGEs or related technologies to spread anti-dengue genes are unlikely to be implemented in the near future as a public health response . Using a biologically detailed model of Aedes aegypti populations dynamics and genetics , we assess how many mosquitoes need to be released to spread an anti-pathogen gene in an urban environment without using an SGE . We compare release numbers with two other , currently feasible transgenic strategies: releasing mosquitoes with female-lethal genes , and mosquitoes carrying both female-lethal and anti-pathogen genes . We show that even without using SGEs , releasing mosquitoes in numbers much smaller than those considered necessary for transgenic population reduction can effectively reduce the ability of Aedes aegypti to spread dengue .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"biotechnology",
"invertebrates",
"ecology",
"and",
"environmental",
"sciences",
"medicine",
"and",
"health",
"sciences",
"tropical",
"diseases",
"animals",
"transgenic",
"engineering",
"population",
"modeling",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"genetic",
"engineering",
"infectious",
"diseases",
"genetically",
"modified",
"organisms",
"theoretical",
"biology",
"epidemiology",
"dengue",
"fever",
"insects",
"disease",
"vectors",
"agriculture",
"arthropoda",
"mosquitoes",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"computational",
"biology",
"evolutionary",
"biology",
"agricultural",
"biotechnology",
"organisms"
] |
2014
|
Feasible Introgression of an Anti-pathogen Transgene into an Urban Mosquito Population without Using Gene-Drive
|
In the late twentieth century , emergence of high rates of treatment failure with antimonial compounds ( SSG ) for visceral leishmaniasis ( VL ) caused a public health crisis in Bihar , India . We hypothesize that exposure to arsenic through drinking contaminated groundwater may be associated with SSG treatment failure due to the development of antimony-resistant parasites . A retrospective cohort design was employed , as antimony treatment is no longer in routine use . The study was performed on patients treated with SSG between 2006 and 2010 . Outcomes of treatment were assessed through a field questionnaire and treatment failure used as a proxy for parasite resistance . Arsenic exposure was quantified through analysis of 5 water samples from within and surrounding the patient’s home . A logistic regression model was used to evaluate the association between arsenic exposure and treatment failure . In a secondary analysis survival curves and Cox regression models were applied to assess the risk of mortality in VL patients exposed to arsenic . One hundred and ten VL patients treated with SSG were analysed . The failure rate with SSG was 59% . Patients with high mean local arsenic level had a non-statistically significant higher risk of treatment failure ( OR = 1 . 78 , 95% CI: 0 . 7–4 . 6 , p = 0 . 23 ) than patients using wells with arsenic concentration <10 μg/L . Twenty one patients died in our cohort , 16 directly as a result of VL . Arsenic levels ≥ 10 μg/L increased the risk of all-cause ( HR 3 . 27; 95% CI: 1 . 4–8 . 1 ) and VL related ( HR 2 . 65; 95% CI: 0 . 96–7 . 65 ) deaths . This was time dependent: 3 months post VL symptom development , elevated risks of all-cause mortality ( HR 8 . 56; 95% CI: 2 . 5–29 . 1 ) and of VL related mortality ( HR 9 . 27; 95% CI: 1 . 8–49 . 0 ) were detected . This study indicates a trend towards increased treatment failure in arsenic exposed patients . The limitations of the retrospective study design may have masked a strong association between arsenic exposure and selection for antimonial resistance in the field . The unanticipated strong correlation between arsenic exposure and VL mortality warrants further investigation .
In the late twentieth century , the emergence of high rates of treatment failure with antimonial compounds for visceral leishmaniasis ( VL ) caused a public health crisis in Bihar state , India [1] . VL , also known as kala-azar , is a significant health problem in India causing up to 282 , 000 clinical cases per year [2] , affecting the most vulnerable populations and leading to significant morbidity , mortality and economic loss [3] . VL is a potentially fatal disease that is caused by the obligate intracellular parasite Leishmania donovani transmitted by the bite of female sand flies of genus Phlebotomus argentipes . In India , VL is diagnosed through visualization of the parasites on splenic or bone marrow biopsy or through rapid diagnostic tests ( e . g . a dipstick detecting antibodies against rK39 ) [4] . Pentavalent antimonial compounds such as sodium stibogluconate ( SSG ) had been the predominant successful treatment against VL during the twentieth century and , in endemic regions such as East Africa , they remain integral and effective in treatment regimens [5 , 6] . In India , since 2005 , SSG use has no longer been officially recommended due to high rates of treatment failure . The VL elimination program in India has been using miltefosine and amphotericin B [7] as an alternative to SSG . However field studies have shown that , if stocks are available , SSG is still prescribed [8] . The reason for the dramatic decline in the efficacy of antimonials in India has previously been attributed to the development of drug resistance in parasites in a context of poor prescribing practices and irrational drug use in India’s under-regulated and largely private health care system [9] . An additional hypothesis has recently been put forward by our group: that background exposure of the population of Bihar through drinking arsenic-contaminated groundwater may have contributed to the development of antimony-resistant parasites [10] . Arsenic and antimony are elements that share many common chemical properties [11] including occurrence in nature in the trivalent and pentavalent states . Leishmania promastigotes resistant to trivalent antimony can be created in the laboratory through stepwise increasing exposure to trivalent arsenic [12] . Naturally occurring arsenic contamination is present in the groundwater accessed by tube wells in many villages of Bihar and , although this was not discovered until 2002 [13] , villagers may have been drinking arsenic contaminated water from tube wells that were first sunk in the region in the late 1970s . It is thus possible that parasites were exposed to accumulated arsenic in the liver and spleen of populations drinking arsenic-contaminated water , and became cross-resistant to antimony through mechanisms developed to deal with arsenic . Proof of concept of this hypothesis has been established in our laboratory by exposing mice to environmentally relevant levels of trivalent arsenic in their drinking water and demonstrating diminished efficacy of the pentavalent antimonial compound SSG in vivo and , in an in vitro macrophage assay , that ex vivo amastigotes from these arsenic exposed mice were resistant to 500 μg/ml SSG [14] . The mode of action of pentavalent antimonial compounds is through activation to their toxic trivalent form resulting in disruption of the thiol metabolism of the parasite [15] , mechanisms which are thought to be disabled in antimonial resistant parasites [9] . Compiling information from published articles , surveys and databases , we identified 10 out of 38 districts in Bihar where VL , arsenic contamination of the groundwater and antimonial treatment failure co-exist [10] . Unfortunately , monitoring of clinical outcomes of VL treatment in this region is sub-optimal [8] and there have been no population-based surveys done here on arsenic contamination so there may be many more districts in Bihar affected by the same conditions . Having established proof of concept of accumulated arsenic inducing pentavalent antimony resistance in vivo in the laboratory , we sought to test the hypothesis that arsenic exposure increases the risk of VL treatment failure with antimonial compounds in the field setting . This presented a number of challenges in study design . First , the study was retrospective , since SSG is no longer widely used in Bihar . Second , it was not possible to test parasite isolates for SSG susceptibility from patients treated years previously . Consequently , treatment failure was used as a proxy for parasite resistance . Although treatment failure can be a multifactorial event , exposure to arsenic varies within the study area . Thus , the primary objective of this study was to evaluate if arsenic exposure is related to treatment failure in SSG-treated patients in Bihar , India . The secondary objective was to evaluate if arsenic exposure is related to death from VL in this antimonial treated cohort .
Ethical clearance was obtained from Banaras Hindu University in Varanasi , the Kala Azar Medical Research Center ( KAMRC ) in Muzaffarpur , India and the University of Dundee , Scotland . Individual written consent was obtained from each patient ( or their guardian ) included in the study . Any person identified with symptoms of VL or post kala-azar dermal leishmaniasis ( PKDL ) during the study was referred to an appropriate health care facility . Data from the arsenic analysis of tube well samples was referred to the Public Health and Education Department , which is running the arsenic mitigation programme in Bihar . The study took place in Mohiuddin Nagar block in the Samastipur district of the state of Bihar , India . Bihar is the third largest state in India and has the lowest literacy rate . Mohiuddin Nagar block lies just north of the River Ganges and has a mainly rural population of 184 , 521 [16] . It is known to be endemic for VL; between 2006–2010 , the average reported yearly incidence of VL was 7 . 78 per 10 , 000 population ( Source: District Malaria Office records , Patna , Bihar ) and Samastipur district , in which it lies , has been mapped as an area with ‘high’ levels of resistance to SSG [17] . A survey performed by the School of Environmental Studies , Jadavpur University in 2005 identified over 40% of the wells surveyed in this district to have arsenic levels above the World Health Organization ( WHO ) recommended limit of 10 μg/L [18] . The fact that SSG treatment for VL is no longer recommended for routine use in Bihar [19] precluded analyzing parasites isolates from patients treated with SSG to determine in vitro resistant phenotypes . To overcome this limitation we designed a retrospective cohort study , using treatment failure as a proxy for presence of parasite resistance . We identified VL patients treated with SSG between 2006 and 2010 in the block of Mohiuddin Nagar from two sources . First , in January and February 2012 , potential SSG treated subjects were identified from the VL patient register at the Primary Health Centre ( PHC ) in Mohiuddin Nagar and were searched for in their listed villages , at least twice , for inclusion in the study . Second , additional VL patients were identified in the visited villages . The patients were included in the study if they gave a clear history of having been treated with SSG ( see below ) and at least one of the SSG treatments fell inside the study recruitment period ( e . g . 2006 to 2010 ) . Patients were excluded if they did not receive SSG treatment or if their SSG treatment was terminated due to unavailability of SSG . The study subjects were visited in their own home . A modified form of a previously validated questionnaire was used [8] to gather information on age , sex , caste , VL symptoms , health seeking behaviour , treatment ( s ) and response from the time of first onset of VL symptoms to the time of the field study . Additional data was collected on other illnesses , prior VL in the family and water sources used ( e . g . local tube wells ) . The interviews were performed by experienced field workers . If the patient was a minor at the time of VL , his/her guardian was interviewed and if the patient had passed away or was unavailable due to relocation for marriage or work , then a relative was interviewed instead . The household tube well and local tube wells were geo-located by the research team using a GPS . The household tube well coordinates were used to identify patients’ location ( e . g . within or outside Mohiuddin Nagar town ) . Clinical records , either the patient’s own documents or those held at the PHC , as well as data gathered during the interviews were reviewed by an experienced physician to ascertain VL cases and treatment . In absence of treatment records , the information provided by patients or relatives allowed identifying the type of treatment received as the 3 main drugs for VL are given via different routes: ( 1 ) SSG is an intramuscular injection usually administered for 30 days in the lateral upper thigh or buttock area; ( 2 ) amphotericin B is given via an intravenous drip in the forearm or hand for either 5 days or 30 days dependent on formulation used and ( 3 ) miltefosine is administered for 28 days as a capsule for oral use . Patients were classified into the following treatment outcome categories: success , no clinical improvement , relapse , death and toxicity taking 6-month cut off time from end of treatment as reference , in accordance with WHO guidance and VL clinical trial protocols [1] . ‘Treatment success’ was defined as patients who received SSG treatment and had not required another VL treatment within 6 months . ‘No clinical improvement’ were patients who experienced no change or a worsening of their original symptoms during or by the end of a full course of treatment who required a further VL treatment . This subgroup included patients who reported a treatment course of 60 days or more which is 2 times the recommended SSG treatment duration . ‘Relapse’ were patients who experienced a return of signs or symptoms of VL , for which they required further treatment , after initially having experienced a return to health following a course of treatment . If the outcome was ‘Death’ , a verbal autopsy was carried out by the physician on the field team to ascertain if the subject died directly as a result of VL . ‘Toxicity’ were patients whose treatment was terminated due to intolerable side effects . The term ‘treatment failure’ in this study covers all adverse outcomes that occurred within 6 months of VL treatment: no clinical improvement , relapse , death due to VL and toxicity . The average arsenic concentration of the patient’s main water source and 4 tube wells surrounding the patient’s home was taken as an indication of the level of arsenic exposure for the patient . This exposure variable was chosen instead of the arsenic concentration in the patient’s primary tube well [20] in the context of this study as many patients reported that their current primary tube well had been inserted in the years after their first SSG treatment for VL . At the time of interview samples were collected from these 5 tube wells in a 15 ml Falcon tube after pumping off water for two minutes . Water samples were preserved with a drop of concentrated nitric acid until analysis . Flow injection hydride generation—atomic absorption spectrometry ( FI-HG-AAS ) was used to quantify the total arsenic content in water samples at the School of Environmental Studies ( SOES ) , Jadvapur University , Calcutta as described previously [13] . A standard sample from the Environmental Protection Agency was used as a reference . The lower detection limit was 3 μg/L of arsenic . For quality control , 25% of the primary tube well water samples were selected to represent the low ( n = 6 ( <10 μg/L ) ) , medium ( n = 15 ( 10–50 μg/L ) ) and high ( n = 9 ( >50 μg/L ) ) arsenic concentrations detected . These samples were re-analyzed by Inductively Coupled Plasma Mass Spectrometry ( ICP-MS ) at Aberdeen University using a method described previously [21] . The average arsenic concentration in the 5 tube wells was dichotomized using the WHO threshold for arsenic ( As ) in water: > = 10 μg/L . The three outcome variables are: ( 1 ) SSG treatment outcome ( primary outcome ) and ( 2 ) all-cause and ( 3 ) VL mortality ( secondary outcomes ) . For the primary analysis , VL patients were classified based on their SSG treatment outcome as “treatment failure” or “treatment success” as described above . For the secondary mortality analyses , the study subjects were classified as “alive” or “dead” at the time of the field study . A sub-group identifying those that died due to VL was created based on the data from the verbal autopsies ( see above ) . Data from the field interviews were double entered in a Microsoft Excel database independently by 2 data entry operators . The covariates included in the analyses are shown in Table 1 . Briefly , age was split into 3 categories ( e . g . < 5 , 6–15 and > 16 years old ) and SSG treated patients were dichotomized based on their “treatment course” using the duration of 30 days recommended by WHO [5] . Patients were classified based on their location ( e . g . in or outside of Mohiuddin Nagar town ) , the “time to treatment” ( e . g . < 12 weeks and ≥ 12 weeks in accordance with previous literature [22] ) and “the place of treatment” ( e . g . government or private facilities ) . Finally , patients were classified into whether they had family members treated for VL with SSG prior to the patients’ VL episode . Dot plots of the log of arsenic exposure were drawn for the outcomes ( 1 ) SSG treatment outcome , ( 2 ) all-cause mortality and ( 3 ) VL mortality and the median arsenic exposures were compared using Mann Whitney U test . Receiver Operating Characteristic ( ROC ) curves were used to evaluate the WHO cut off for arsenic in water ( 10 μg/L ) against the primary and secondary outcomes ( e . g . SSG treatment , all-cause and VL mortality ) . The Kappa index was used to evaluate the agreement between the arsenic levels reported by the two laboratories ( SOES and Aberdeen University ) that analyzed the quality control samples . A logistic regression model was employed to assess if arsenic contamination in the local environment increased the risk of treatment failure in SSG treated patients . First , each covariate was analyzed against the main outcome using logistic regression . Then , a multivariate logistic regression model was built , using a forwards stepwise method , with three a priori selected variables ( “forced” variables: age , sex and location ) . Any additional variables which had a p value of <0 . 2 in the bivariate analyses were evaluated in the multivariate model and retained if they had a p value of < 0 . 05 . The results were presented as odds ratios ( OR ) and their 95% Confidence Intervals ( CI ) . Survival analysis methods were used to compare survival in subjects exposed to elevated ( > = 10 μg/L ) versus normal arsenic water levels . The time origin for the survival analysis was the reported start date of symptoms . The date of 01/02/12 , mid date of field visits , was used as the censor date . First , a Kaplan Meier ( KM ) survival curve and the log-rank test were used to evaluate the association between arsenic in water and all-cause mortality . Cox regression was used to estimate the risk of all-cause mortality in patients exposed to arsenic contaminated wells while controlling for possible confounding factors . The proportional hazards assumption was tested using visual methods and on the basis of scaled Schoenfield residuals [23] . If the proportional hazard assumption was invalid the effect of arsenic exposure was allowed to vary over time by fitting 2 time varying covariates in the model . The KM plots were used to determine the temporal periods in which the effect of arsenic changed . The final regression model included arsenic exposure , three a priori selected variables ( “forced” variables: age , sex and location ) and any variables that were associated with mortality as assessed by the log rank test ( p <0 . 2 ) and remained significant ( p <0 . 05 ) in the final model . Hazard ratios ( HR ) and their 95% CI were used to express the risk in the final model . Analogous analyses ( e . g . KM and Cox regression model ) were used to study the risk of VL mortality associated with arsenic exposure .
We identified 606 patients treated for VL from January 1st 2006 to December 31st 2010 at the PHC of Mohiuddin Nagar . One hundred and thirty four ( 22% ) of them were treated with SSG . The rest ( n = 472 ) were excluded from the study as they received miltefosine or amphotericin B ( n = 470 ) or were duplicate entries ( n = 2 ) . Twenty-five additional VL patients treated with SSG from 2006 to 2010 were identified at the time of the fieldwork visits . Out of the 159 SSG treated patients identified , the house of 113 ( 71% ) of them was located . The reasons for being unable to locate patients included incomplete/inaccurate information ( n = 43 ) and distance from Mohiuddin Nagar block ( n = 3 ) . From the 113 patients , 69 ( 61% ) of them were present for interview . Twenty one ( 19% ) of the subjects were dead and 23 ( 20% ) were living outside the study area at the time of the visit . Relatives from the 44 missing subjects were interviewed . Based on the information gathered , 3 patients were excluded as they had their SSG treatment terminated due to unavailability of the drug . A cohort of 110 VL patients treated with SSG was finally included in the analyses ( Fig . 1 ) . The 110 patients were aged between 3 and 60 years old , with a median age of 14 and a ratio of males to females of 3:2 . Twenty percent of subjects lived within the area of Mohiuddin Nagar town ( Fig . 2 ) . Twenty-three ( 21% ) of patients had experienced other illnesses prior to their VL episode including the infectious diseases of malaria ( n = 2 ) , tuberculosis ( n = 2 ) , hepatitis/jaundice ( n = 4 ) , cholera ( n = 1 ) and polio ( n = 1 ) . Non-infectious illnesses included asthma ( n = 7 ) , diabetes ( n = 1 ) , arthritis ( n = 1 ) , eczema ( n = 1 ) and neurological complaints ( n = 3 ) . All study subjects reported similar symptoms from their VL episode of fever , anorexia and weight loss ( 100% , 91% , 90% respectively ) with malaise , abdominal distension and skin pigmentation being less common ( 72% , 44% and 7% respectively ) . Confirmation of the diagnosis of VL was available in the form of record of a positive rapid diagnostic test for VL ( e . g . rK39 dipstick ) in 26 subjects ( 24% ) . Parasitological confirmation was not available for any patient . Seventy-one patients ( 65% ) were only treated with SSG; 52 received one treatment , 18 received two and 1 patient received three SSG courses . Fourteen out of the 52 patients who received one treatment reported SSG treatment courses of 60 and 90 days duration which are 2 and 3 times the recommended duration . Thirty-nine patients ( 35% ) were treated with miltefosine or amphotericin B before ( n = 3 ) or after ( n = 34 ) SSG treatment or both ( n = 2 ) . The mean number of treatments ( using any drug ) received per person was 1 . 6 ( ±0 . 6 ) . The number of clinical documents confirming the treatments administered was limited . For example , only 16 treatment records were available for review out of the 85 patients treated at the PHC . According to those records , the treatment of 7 out of 16 patients ( 44% ) did not conform to WHO recommendations—either by dose ( only 10mg/kg ) or duration ( 20 vs 30d ) . However , the treatment duration reported by the patient in the questionnaire matched the PHC treatment card only in 10 out of these 16 patients ( 63% ) . Information on doses was not available from the patients’ interviews . Sixty two ( 56% ) of the study subjects were classified as “treatment failure”: 10 ( 9% ) died due to VL within 6 months , 40 ( 36% ) patients experienced no clinical improvement , 8 ( 7% ) patients relapsed and 4 ( 4% ) terminated their treatment due to toxicity ( Fig . 1 ) . There was a high death rate in our cohort: 21 ( 19% ) of 110 patients had died by the date of the interview . Sixteen ( 15% ) of these patients died from VL , 6 of these deaths occurred after the 6 month follow up period for SSG treatment failure so they are not included in the primary outcome . One of these deaths was directly due to treatment toxicity . Five patients died from non-VL causes including HIV ( n = 1 ) , asthma ( n = 1 ) , liver failure ( n = 1 ) , paralysis ( n = 1 ) and a road traffic accident ( n = 1 ) . Treatment outcomes were not recorded in the PHC register , PHC treatment cards or seldom in patient’s hand held documents . The arsenic concentration in the wells tested ranged from < 3 μg/L up to 1050 μg/L arsenic . Fifty ( 44% ) of the study subjects had at least one arsenic contaminated tube well in their home or surrounding area and 26 ( 24% ) of them had a mean local arsenic level ≥ 10 μg/L ( Table 1 and Fig . 2 ) . Five ( 4 . 5% ) patients had a mean local arsenic level of >50 μg/L , the previous Indian Government limit . Only 11 ( 10% ) patients were aware of the issue of arsenic contamination and 8 of these lived in an area with local arsenic contamination . The remaining 18 out of the 26 patients living with local arsenic contamination had no knowledge of the issue . For the cut off ≥10 μg/L arsenic the areas under the ROC curves were ( 1 ) 0 . 54 for treatment outcome , ( 2 ) 0 . 69 for all-cause mortality and ( 3 ) 0 . 65 for VL mortality . The sum of sensitivity and specificity was greater than 100% for all outcomes . The quality control assays showed a good agreement ( Kappa = 90% , p<0 . 001 ) between the analysis performed at SOES and the University of Aberdeen laboratories . Although the median arsenic water exposure level was higher in patients with treatment failure ( Fig . 3 , panel A ) , this difference was not statistically significant ( Mann Whitney U p = 0 . 42 ) . Only one covariate had a weak association with treatment failure ( p<0 . 2 ) in the bivariate analysis: previous SSG treatment in the family ( p = 0 . 11 ) . This variable was not retained in the final multivariate model for lack of significance . Although the association between treatment duration and outcome was statistically significant ( p = 0 . 01 ) this variable was not included in the final model as in the majority of cases the treatment duration is dependent on the outcome . The final logistic regression model only included the forced variables age , sex and location and shows a trend: patients with high mean local arsenic level have a higher risk of treatment failure ( OR = 1 . 78 ) than patients using wells with arsenic concentration <10 μg/L , however this association is not statistically significant ( 95% CI: 0 . 7–4 . 6 , p = 0 . 23 ) . As shown in the dot plots ( Fig . 3 , panels B and C ) the median arsenic concentrations in water are significantly higher in the patients who died ( group “all-cause mortality” ( p = 0 . 005 ) and those who died due to VL ( group “VL mortality” p = 0 . 048 ) compared to the patients that were still alive at the time of the field visits . The KM curves ( Fig . 4 , panels A and B ) confirmed increased mortality in subjects exposed to elevated arsenic water levels ( p = 0 . 004 and p = 0 . 044 for “all-cause” and “VL mortality , ” respectively ) . Age was the only covariate to show a significant association with mortality and thus the final models only included the forced variables of age , sex and location . The multivariate Cox regression model showed that arsenic levels ≥ 10 μg/L increased significantly the risk of all-cause ( HR 3 . 27; 95% CI: 1 . 4–8 . 1 ) and VL related ( HR 2 . 65; 95% CI: 0 . 96–7 . 65 ) mortalities . However , the effect of arsenic exposure on all-cause mortality was found to vary with time . On the KM plots , the two survival lines representing arsenic exposed and non-exposed populations separate at 3 months where the effect of arsenic exposure becomes apparent . This was confirmed by a modified Cox regression model which allowed the effect of arsenic exposure to vary over time . No increased risk of mortality was detected with arsenic exposure during the first 3 months after start of VL symptoms . However , after 3 months , the effect of arsenic exposure significantly increased the risk of all-cause mortality ( HR 8 . 56; 95% CI: 2 . 5–29 . 1 ) and of VL related mortality ( HR 9 . 27; 95% CI: 1 . 8–49 . 0 ) ( Table 2 ) .
The results of this study found no significant relationship between exposure to arsenic and SSG treatment failure . However there was an unexpectedly high mortality rate in the cohort of SSG treated patients and a secondary analysis of mortality showed that arsenic exposure strongly increased the risk of both all-cause mortality and VL mortality . The primary objective of this study was to evaluate if arsenic exposure is related to treatment failure in SSG treated patients . A positive correlation would have supported the laboratory finding that SSG resistant Leishmania parasites can be generated in a mouse model of oral arsenic exposure at environmentally relevant levels [14] . The results from this study , although they indicate a trend towards increased treatment failure in arsenic exposed patients , fail to confirm that this resistance generation mechanism is operating in the field . Analogous findings were obtained when the SSG treatment outcome was evaluated against the individual levels of arsenic in urine in the subset of patients where these were available ( S1 Fig . and S1 Table in S1 Text ) . A number of reasons may explain why the field results do not support the laboratory results . Firstly , the numbers of patients available were limited as SSG treatment for VL is no longer recommended in Bihar and the study was underpowered to detect a low level of risk . Secondly , the in vivo experiments were performed in the upper range of arsenic levels found in Bihar . Only 2 patients in our study had arsenic levels in their local water supplies in this range and they both had a successful treatment with SSG . Thirdly , treatment failure from SSG is thought to be multifactorial and not only due to resistant parasites , particularly in that there is low correlation between in vitro resistance tests and clinical outcome [24] . Additionally , extensive work on SSG resistant parasites has shown that they have increased fitness and virulence when compared to SSG sensitive strains indicating that they would thus be preferentially transmitted and high levels of resistant parasites could be circulating [25] . These latter two reasons could mask any existing relationship between arsenic exposure and SSG treatment failure in VL . Although multiple studies have been carried out on the mechanisms of antimonial resistance in the Leishmania parasite [9] , only a few epidemiological studies have been performed looking at the clinical and demographic risk factors associated with SSG treatment failure in the leishmaniases [22 , 26] . The most relevant is a prospective study in Nepal on VL patients [22] that identified patients living on the border of Bihar as having a markedly increased chance of treatment failure . Additionally , fever over 12 weeks , interruption of treatment and ambulatory treatment were strong risk factors for treatment failure . All of the patients in our study received ambulatory treatment , treatment interruption was not specifically assessed and no association was found between prolonged fever and treatment failure . As arsenic contamination is an issue also in Nepal , it would be interesting to assess retrospectively the arsenic exposure in this cohort of patients . A study on rural VL care in Muzaffarpur district , Bihar [8] ( northwest of the study area box in Fig . 2B ) where no significant arsenic contamination has been detected [10] , but SSG resistance is well established [1] , had a SSG failure rate of 40% compared with 59% in our study . The difference in failure rates in these similar communities , with comparable district level literacy rates of 61 . 9% and 63 . 5% respectively [16] may be , among other factors , attributable to arsenic exposure and the increased mortality rate . The main finding from our study , the strong relationship found between arsenic exposure and all-cause and VL mortality agrees with a population-based cohort study performed in Bangladesh [20] that demonstrated an increased risk for both all-cause mortality and infectious disease deaths with increasing arsenic exposure . The risk of death from VL in arsenic-exposed persons in our cohort is predominantly present greater than 3 months post the commencement of symptoms . This could be explained by 3 mechanisms: 1 ) the longer incubation time in arsenic exposed tissues would allow the parasites to generate arsenic tolerance and thus SSG resistance leading to ineffective treatment; 2 ) patients being infected with SSG-resistant parasites [17]; and 3 ) the risk of mortality increases with the progressive dampening effect of VL on the immune system [27] combined with the immunotoxicity of arsenic exposure [28] . The rural VL study in Muzaffarpur referred to above [8] only reported one death ( 0 . 7% ) of undefined cause in a cohort of 138 ( 68 of whom were treated with SSG ) compared with 21 deaths ( 19% ) in our 110 patient cohort with 16 directly as a result of VL ( 14 . 5% ) . This could be explained by a number of factors as well as arsenic exposure such as: a different quality of primary health care between districts; a more virulent parasite population; or a general lower immune status , for example due to malnutrition , in our Samastipur study cohort . However , if the immunotoxicity of arsenic and VL combined is responsible for the increased death rate in our study then arsenic exposure may also increase mortality from VL with treatments other than SSG . This warrants further research . Previous studies on mortality in VL have identified the extremes of age , ( <5 and > 45 years old ) , long duration of symptoms , co-infections and laboratory abnormalities such as severe anaemia and jaundice as risk factors for death [29–32] . Due to the retrospective design of our study , and the rural management of the VL cases , information on the above clinical risk factors is mainly not available . Our data does agree with age associated risk at the extremes of age but no increased mortality risk was seen with delay to first SSG treatment . It would have been ideal to perform a prospective study where the parasites’ sensitivity to pentavalent antimony in vitro in macrophages was correlated with arsenic exposure and the clinical outcome of SSG treatment . However , due to the high rate of treatment failure with antimony and the recommendation of discontinuation of use , this type of prospective study was not possible . The retrospective nature of the study meant that parasite isolates from antimony-treated cases were not available . The presence of SSG-resistant parasites was reported earlier in the Samastipur District [33] . Determining if antimonial resistant parasites are still circulating in the study area today may have helped to interpret the findings of our study , provided that stably resistant parasites had not extensively displaced SSG-sensitive isolates due to their increased fitness [34] . The retrospective design of this study also had some limitations in ascertaining VL cases and treatments used . No parasitological confirmation was available but most of the patients were diagnosed with VL by the rK39 rapid diagnostic test which shows a good specificity ( 90 . 6% ) in the context of clinically suspected disease [35] . Relying on patient and relatives responses increased the risk of recall bias [36] . Therefore , the definition of treatment failure included the requirement for another treatment course which is more concrete that a subjective analysis of symptoms . As SSG treatment is prolonged and painful with a dramatic economic impact on the patient’s family , the recall bias for treatment may be limited . However , it is possible that administered doses were too low , or treatments were given with unrecalled interruptions that could have impacted on treatment outcome . Unfortunately , there was generally poor documentation of diagnostic methods , treatment duration and dosing . Finally , VL as a cause of death was confirmed by verbal autopsy only carried out by one physician . For this reason VL mortality is always presented with all-cause mortality . A further substantial design difficulty in this study is the timing of the assessment of arsenic exposure . Forty percent of the patients had their tube well inserted in the years following their treatment episode and therefore a proxy of their arsenic exposure had to be created from the average of 5 wells surrounding their living area . Additionally , the level of arsenic contamination in a tube well water sample can vary according to prior usage that day , depth of well and age of well [37 , 38] adding further factors into the difficulty of building up an accurate picture of arsenic exposure at a historic time point . Although water arsenic levels were being collected on average 5 ( ± 1 . 3 ) years post treatment , it is generally assumed that arsenic in tube wells in the Bengal basin is either stable over time [20 , 39–41] or may rise gradually [37 , 38] . This cohort study was performed in an area where research on VL has not been undertaken previously and gives a view of practice in an area without any special intervention . It highlights the issue , previously identified in Muzaffarpur [8] , that SSG was still being used in just under a quarter of patients treated between 2006 and 2010 , with no record of its generally poor outcome , despite recommendations for its discontinuation since 2005 [42] . In the registers available at the PHC there was no indication of any treatment failure or relapse and , on interview , the staff were only aware of one death during the study period . The majority of patients ( 57% ) felt the need to change health care provider to obtain an effective treatment . The doctors at the PHC knew of the issue of SSG resistance and have been aware of a drive to mainly use miltefosine or amphotericin but , worryingly , were unaware of the VL elimination programme . They were aware of the issue of arsenic contamination but unaware of its wide-ranging health effects . This work shows the urgent need for improved record keeping , education and intervention in both VL and arsenic contamination within these vulnerable communities [3 , 20] . A recent global review of the epidemiology of the leishmaniases gave an overall case fatality rate for VL of 10% [2] which is considerably lower than the VL mortality rate of 14 . 5% in our cohort . The study demonstrates that arsenic exposure is strongly associated with VL mortality in this antimony treated cohort and may be responsible for the elevated mortality rate . Unfortunately , the study was underpowered to confirm anything but a strong association between arsenic exposure and SSG treatment failure . This research into antimonial resistance and treatment failure has important implications for the leishmaniases worldwide and is a reminder to consider the environment in which an organism is propagating when assessing reasons for treatment failure and mortality . Further research into the relationship between arsenic exposure and VL mortality is required .
|
The parasitic disease visceral leishmaniasis ( VL ) causes a significant burden of illness and death in India . The main drug used to treat VL , which is based on the chemical element antimony , stopped working well in about half of all patients in the late twentieth century . We hypothesised that arsenic exposure of the Indian population , through contaminated groundwater , was contributing to treatment failure with antimony based drugs . Arsenic and antimony are similar chemical elements and exposure of the parasite to arsenic within the liver of arsenic-exposed patients could allow the parasite to become resistant to treatment with antimony . Using a field-based questionnaire study we retrospectively evaluated whether arsenic exposure was linked to antimonial treatment failure in a cohort of 110 antimonial treated patients . No significant association was found , although this may be because the number of patients in the study was low as antimony use was officially discontinued in 2005 due to high rates of treatment failure . However , arsenic exposure was found to increase risk of mortality from VL particularly if death occurred more than 3 months after the symptoms of VL developed . More research into the relationship between arsenic exposure and mortality in VL is warranted .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Arsenic Exposure and Outcomes of Antimonial Treatment in Visceral Leishmaniasis Patients in Bihar, India: A Retrospective Cohort Study
|
The reduced amounts of Mycobacterium leprae ( M . leprae ) among paucibacillary ( PB ) patients reflect the need to further optimize methods for leprosy diagnosis . An increasing number of reports have shown that droplet digital polymerase chain reaction ( ddPCR ) is a promising tool for diagnosis of infectious disease among samples with low copy number . To date , no publications have investigated the utility of ddPCR in the detection of M . leprae . The aim of this study was to develop and evaluate a ddPCR assay for the diagnosis of PB leprosy . The two most sensitive DNA targets for detection of M . leprae were selected from electronic databases for assessment of sensitivity and specificity by quantitative polymerase chain reaction ( qPCR ) and ddPCR . Control patients ( n = 59 ) suffering from other dermatological diseases were used to define the cut-off of the duplex ddPCR assay . For comparative evaluation , qPCR and ddPCR assays were performed in 44 PB patients and 68 multibacillary ( MB ) patients . M . leprae-specific repetitive element ( RLEP ) and groEL ( encoding the 65 kDa molecular chaperone GroEL ) were used to develop the ddPCR assay by systematically analyzing specificity and sensitivity . Based on the defined cut-off value , the ddPCR assay showed greater sensitivity in detecting M . leprae DNA in PB patients compared with qPCR ( 79 . 5% vs 36 . 4% ) , while both assays have a 100% sensitivity in MB patients . We developed and evaluated a duplex ddPCR assay for leprosy diagnosis in skin biopsy samples from leprosy patients . While still costly , ddPCR might be a promising diagnostic tool for detection of PB leprosy .
Leprosy , a chronic infectious disease caused by M . leprae , has a tropism for macrophages in skin and Schwann cells in peripheral nerves [1] . This disease is quite variable , affecting people in different ways according to their immune response . At one end of the spectrum , patients with a high level of immunity harbor a low number of bacilli and are termed PB patients . Patients with many bacilli are referred to as MB patients [2] . Despite its elimination as a global public health problem due to the widespread implementation of multidrug therapy , leprosy continues to mar the lives of the infected individuals [3] . In 2016 , a total of 214 , 783 new patients of which 12 , 819 were detected with visible deformities , were reported in 143 countries among all World Health Organization regions filed , corresponding to a global new case detection rate of 2 . 9 per 100 , 000 population[4] . The principal consideration in measuring the reduction of leprosy burden has been the decrease the number of visible deformities among new patients [4] . Therefore , early diagnosis and prompt treatment remain key strategies for leprosy control [1 , 4] . Because the main diagnostic tools for leprosy involve bacillary counts with a limited sensitivity of 30% and histopathology showing a specific neural inflammation histopathologic changes , which require well-experienced clinicians , late diagnosis is frequently the case for many patients [2 , 3] . Although serological tests and IFN-γ releasing assays have also been used to detect M . leprae as potential diagnostic tools [5 , 6] , PB patients are negative for phenolic glycolipid-1 and household contacts exhibit a similar pattern of IFN-γ secretion as PB patients [5–7] . In the past three decades , identification of M . leprae DNA has become popular through the development of PCR methods for leprosy diagnosis [4 , 7] . As 33%-83% of PB patients have been confirmed by PCR , this has greatly aided clinicians in identifying leprosy patients with negative bacilloscopic and inconclusive histopathological features [7] . For MB patients who have high bacillary loads are easily detected by PCR , and the sensitivity of qPCR is almost 100% [7] . The ddPCR , based on water-oil emulsion droplet technology , is a new PCR method for nucleic acid detection [8–11] . Several studies on ddPCR have shown its higher sensitivity and precision in molecular diagnostics for pathogens such as hepatitis B virus [8] , human immunodeficiency virus ( HIV ) [9] , chlamydia trachomatis [10] and chromosomally integrated human herpes virus 6 [11] . To the best of our knowledge , no publications have reported on the clinical utility of the ddPCR assay for leprosy . Here , we developed a ddPCR assay for the diagnosis of leprosy in skin biopsy specimens and compared the diagnostic performance of ddPCR and qPCR on leprosy .
The study was approved by the institutional review board ( IRB ) committee of the Shandong Provincial Institute of Dermatology and Venereology , Shandong Academy of Medical Science , China ( IRB approval number: 2016-KYKT-29 ) . We followed the Genetic Risk Prediction Studies guidelines [12] and written informed consent was obtained from each participant and all of whom were adult subjects . A total of 112 leprosy patients ( comprising 68 MB and 44 PB patients ) and 59 non-leprosy patients from Shandong Provincial Hospital for Skin Diseases ( Shandong , China ) were collected and enrolled in this study . All patients were of Chinese descent . The confirmed diagnoses were based on systematic analysis and integration of patients’ medical history , clinical manifestations , slit skin smear staining , histological examinations . We used Mycobacterium marinum ( M . marinum ) and Mycobacterium tuberculosis to evaluate the specificity of the assays . M . marinum was provided by Dr . Annemarie H . Meijer ( Department of Molecular Cell Biology , Institute of Biology , Leiden University , Leiden , Netherlands ) and eight DNA samples from sputum of patients infected by Mycobacterium tuberculosis were provided by Jinan infectious disease hospital . DNA was extracted from skin biopsies and M . marinum using QIAamp DNA Mini Kits ( Qiagen ) according to the manufacturer’s instructions . Extracted DNA was measured with a NanoDrop 8000 spectrophotometer ( Thermo Scientific ) and then either used immediately or stored at -80°C . Following the guidelines for reporting systematic reviews from PRISMA [13] , we searched PubMed and EMBASE from their inception until March 25 , 2018 to assure a comprehensive study . Six genes , including RLEP , 18 kDa heat shock protein ( HSP18 ) , antigen 85B ( Ag 85B ) , superoxide dismutase A ( sodA ) , 16S ribosomal Ribose Nucleic Acid ( 16SrRNA ) and early secretory antigenic target ( esxA ) , have been used in Taqman qPCR previously [14–17] . For other 11 genes , their primers and probes were designed by Premier 3 . 0 based on the DNA sequences in previous studies [17–26] . The primers and probes of all 17 genes were summarized in S1 Table . Five DNA samples were chosen among 68 MB patients as representative to systematically evaluate the sensitivity of 17 genes . Briefly , the DNA was firstly normalized using ddPCR based on the target gene of Ag85B , which had shown as the most specific target gene in previous publications [14] . Then the DNA samples were diluted to 1 , 000 copies/ul , followed by increasing dilutions ( 1:10 , 1:100 , 1:200 , 1:1 , 000 , 1:2 , 000 , 1:10 , 000 , 1:20 , 000 and 1:100 , 000 ) . Finally , the two most sensitive genes ( RLEP and groEL ) from 17 target genes were selected according to the highest dilutions that could be detected by qPCR and ddPCR ( limit of detection ( LOD ) ) . qPCR was performed in duplicate using the ABI Step One Plus real-time PCR system ( Applied BioSystems ) . PCR reaction mixtures were 20 μL in volume and contained 10 μL of 2× TaqMan Gene Expression Master Mix ( Applied BioSystems ) , 900 nM primers , 250 nM probes and 4 μL of extracted DNA . The qPCR condition was as follows: 50˚C for 2 min and 95˚C for 10 min , followed by 40 cycles of 15 s at 95˚C and 1 min at 60˚C . Fluorescent accumulation data were analyzed using the ABI StepOne Software Version 2 . 2 . 2 ( Applied Biosystems ) . The threshold cycle ( CT ) values of < 37 was defined a positive result for the qPCR assay . After determination of the two most sensitive target genes ( RLEP and groEL ) , the qPCR was performed in all samples enrolled in this study , which were considered as positive when three or four wells ( RLEP and groEL in duplicate ) have positive signals ( CT< 37 ) . The ddPCR was performed in duplicate using a QX200 Droplet Digital PCR system ( Bio-Rad ) . Each assay mix was prepared in a final volume of 20 μL , containing 10 μL of 2× ddPCR Supermix for Probes ( no dUTP; Bio-Rad ) , 900 nM primers , 250 nM probes and 4 μL of extracted DNA . The generation of droplets was performed by the QX200 Droplet Generator ( Bio-Rad ) according to the manufacturer’s protocols . PCR amplification was carried out on an Applied Biosystems Veriti 96-Well Thermal Cycler using the following PCR conditions: 95°C for 10 min followed by 40 cycles of 94°C for 30 s , 60°C for 1 min and a final extension step at 98°C for 5 min . The plate was stored at 16°C until droplets were analyzed by the QX200 Droplet Reader and QuantaSoft software version 1 . 7 . 4 ( Bio-Rad ) . The ddPCR of RLEP and groEL genes was performed in all samples enrolled in this study , and the fluorescent signal events above the threshold line were evaluated . A positive well was defined if more than four fluorescent signal events were shown above the threshold line . The samples were determined as positive when the four test wells ( RLEP and groEL in duplicate ) showed at least three positive wells . The detailed protocols regarding qPCR and ddPCR are available in protocols . io in the following: dx . doi . org/10 . 17504/protocols . io . v4ye8xw; dx . doi . org/10 . 17504/protocols . io . v4ze8x6 . Data were statistically described in terms of range , mean ± standard deviation ( SD ) , frequency ( number of patients ) and relative frequency ( percentages ) . The statistical significance of the differences in sensitivities between ddPCR and qPCR were assessed by means of the kappa test and McNemar test . The differences of age between MB , PB and non-leprosy patients were assessed by ANOVA test , race and gender were assessed by Chi-square test . This manuscript followed the Standards for the Reporting of Diagnostic accuracy studies ( STARD ) ( S1 File , S2 File , S3 File ) .
A total of 171 patients including 68 MB patients , 44 PB patients and 59 non-leprosy patients were enrolled in this study . All clinical characteristics of these 171 subjects are provided in Table 1 . This study consisted of 109 males ( 63 . 7% ) and 62 females ( 36 . 3% ) . The mean age of MB patients , PB patients and non-leprosy patients were 44 . 9 ( range from 13 to 77 ) , 45 . 7 ( range from 19 to 80 ) and 44 . 8 ( range from 18 to 78 ) , respectively . 135 subjects ( 78 . 9% ) were Chinese Han descent . There was no difference regarding the gender , age and ethnicity among these three groups ( all P values > 0 . 05 ) . Every primers and their related probes of all 17 genes were aligned with the M . leprae genome using Basic Local Alignment Search Tool software ( NCBI ) . The results of the sensitivities of 17 target genes detected by qPCR in five MB patients are shown in S2 Table . RLEP , groEL , proline-rich antigen ( pra ) , esxA , HSP18 and 85B target genes showed a higher sensitivity than other genes , of which LOD was lower than 1:2 , 000 ( 0 . 5 copies/ul ) . Among the six target genes , the two most sensitive genes were RLEP and groEL , given that the CT values of RLEP and groEL were less than 38 at the dilution ratios of 1:2 , 000 . Moreover , RLEP and groEL showed more than 10 times the sensitivity of the other four genes in ddPCR ( Fig 1 ) . We further evaluated the specificity of RLEP and groEL genes . Neither Mycobacterium tuberculosis nor M . marinum yielded positive results by qPCR or ddPCR . Therefore , RLEP and groEL were finally selected as the DNA targets to establish the ddPCR assay . Skin biopsies from 59 non-leprosy patients that were diagnosed as inflammatory diseases , such as psoriasis , lichen planus , served as negative controls to define the cut-off of the duplex ddPCR assay . The mean positive events for RLEP were 0 . 34±0 . 56 ( 95% CI 0 . 19–0 . 49 ) and the maximum value was two . For groEL , the mean positive events were 0 . 49±0 . 67 ( 95% CI 0 . 32–0 . 67 ) with a maximum score of three ( S3 Table ) . A positive result of ddPCR assay was determined as follows: 1 ) the threshold line for RLEP and groEL was 5 , 000 and 2 , 500 , respectively ( Fig 1 ) ; 2 ) the well was marked as a positive well if more than four fluorescent signal events were shown above the threshold line to avoid false positive; and 3 ) the sample , which was present in at least three positive wells , was defined as an M . leprae-infected sample . Of the 68 MB patients , the sensitivity of qPCR and ddPCR were both 100% . No case of non-leprosy patients showed positive results in both qPCR and ddPCR assays , showing a specificity of 100% . Out of 44 PB patients , qPCR was positive in 16 patients ( 36 . 4%; 95% confidence interval [CI] , 23 . 7 to 51 . 2% ) . In contrast , ddPCR detected M . leprae in 35 patients ( 79 . 5%; 95% CI , 65 . 3 to 89 . 1% ) . A total of 16 patients ( 36 . 4%; 95% CI , 23 . 7 to 51 . 2% ) tested positively by both qPCR and ddPCR . There was no case in which qPCR was positive and ddPCR was negative . The ddPCR confirmed the diagnosis in 19 out of 28 skin tissues ( 67 . 9%; 95% CI , 49 . 2 to 82 . 2% ) which were qPCR negative ( Table 2 ) . Comparative analysis of the positivity between qPCR and ddPCR indicated that the sensitivity of ddPCR was significantly higher than that of qPCR in our study ( P<0 . 001 ) .
RLEP: NC_002677 . 1 ( 39269 . 39991 ) . groEL: Gene ID: 908906 . pra: Gene ID: 908610 . esxA: Gene ID: 908212 . HSP18: Gene ID: 910696 . 85B: Gene ID: 909036 . rpoT: Gene ID: 910077 . ML0024: Gene ID: 909040 . ML1545: Gene ID: 909602 . ML2179: Gene ID: 908978 . soda: Gene ID: 910514 . 16SrRNA: Gene ID: 910245 . TTC: Gene ID: 908674/908673 . ML0098: Gene ID: 908293 . AT: Gene ID: 909755/909757 . MntH: Gene ID: 908932 . AGT: Gene ID: 908866/908865 .
|
Leprosy , or Hansen’s disease , is a chronic bacterial disease caused by M . leprae . Although it is curable and early treatment averts most disabilities , it remains an important global health concern . This is mainly due to delayed diagnosis . In leprosy , a reliable and early diagnostic tool , is still needed . In recent decades years , the quantitative PCR ( qPCR ) based on nucleic acid detection has been employed for leprosy diagnosis , which exhibited high sensitivity . The performance of qPCR assays , however , greatly varied in different studies , especially in the diagnosis of PB patients . ddPCR is a new and sensitive method used in the examination of pathogenic microorganism , showing considerable reliability and efficiency in other infectious diseases . To our knowledge , no publication reported the ddPCR assay for leprosy diagnosis . Herein , we developed and evaluated a ddPCR assay for detecting M . leprae in skin biopsy samples . Our results suggest that ddPCR specially targeting RLEP and groEL genes could be a promising tool to the detection of M . leprae in PB leprosy with a higher sensitivity than qPCR . This research provides a new molecular biology methods for leprosy diagnosis .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"mycobacterium",
"leprae",
"medicine",
"and",
"health",
"sciences",
"biopsy",
"tropical",
"diseases",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"organisms",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"dna",
"bacteria",
"extraction",
"techniques",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"artificial",
"gene",
"amplification",
"and",
"extension",
"actinobacteria",
"molecular",
"biology",
"biochemistry",
"dna",
"extraction",
"diagnostic",
"medicine",
"nucleic",
"acids",
"genetics",
"leprosy",
"biology",
"and",
"life",
"sciences",
"polymerase",
"chain",
"reaction"
] |
2019
|
Development and evaluation of a droplet digital PCR assay for the diagnosis of paucibacillary leprosy in skin biopsy specimens
|
Certain microbes invade brain microvascular endothelial cells ( BMECs ) to breach the blood-brain barrier ( BBB ) and establish central nervous system ( CNS ) infection . Here we use the leading meningitis pathogen group B Streptococcus ( GBS ) together with insect and mammalian infection models to probe a potential role of glycosaminoglycan ( GAG ) interactions in the pathogenesis of CNS entry . Site-directed mutagenesis of a GAG-binding domain of the surface GBS alpha C protein impeded GBS penetration of the Drosophila BBB in vivo and diminished GBS adherence to and invasion of human BMECs in vitro . Conversely , genetic impairment of GAG expression in flies or mice reduced GBS dissemination into the brain . These complementary approaches identify a role for bacterial-GAG interactions in the pathogenesis of CNS infection . Our results also highlight how the simpler yet genetically conserved Drosophila GAG pathways can provide a model organism to screen candidate molecules that can interrupt pathogen-GAG interactions for future therapeutic applications .
Bacterial meningitis is one of the top ten causes of infection-related mortality worldwide [1] . Meningitis is particularly devastating in the newborn infant , and 20–50% of survivors can suffer permanent neurological sequelae including deafness , seizures , hydrocephalus , cerebral palsy and/or cognitive deficits [2]–[5] . The most common agent of neonatal bacterial meningitis in the United States , Europe and Asia is group B Streptococcus ( GBS ) . In recent years , GBS has also emerged as a cause of serious infections including meningitis in nonpregnant adult populations , with an invasive disease incidence approaching that reported for the neonate [6]–[8] . In gaining access to the central nervous system ( CNS ) , GBS reveals an ability to cross the blood-brain barrier ( BBB ) , a specialized layer of brain microvascular endothelial cells ( BMECs ) that regulates macromolecular traffic to maintain biochemical homeostasis in brain tissues . BBB penetration by a bacterial pathogen reflects a complex interplay between host endothelium and microbial products [9] . The fundamental mechanisms by which GBS establishes CNS infection remain incompletely understood [10] . The plasma membrane of mammalian endothelial cells is prominently decorated with linear , negatively charged sugar chains known as glycosaminoglycans ( GAGs ) . The sulfated GAGs , chondroitin/dermatan sulfate and heparan sulfate , occur as proteoglycans that consist of one or more GAG chains covalently linked to a core protein . Cells elaborate several membrane heparan sulfate proteoglycans ( HSPGs ) including syndecans and the glycosylphosphatidylinositol-linked glypicans . The heparan sulfate chains assemble by copolymerization of alternating residues of N-acetylglucosamine and glucuronic acid , which then can undergo variable sulfation and uronic acid epimerization . This process occurs in a non-template driven manner , resulting in sections of the chains with variably modified sugars of variable length interspersed with domains of unmodified residues . The modified regions make up binding sites for growth factors , membrane receptors , and extracellular matrix constituents , imparting to the chains important biological activities , such as roles in cell adhesion , receptor signaling , and leukocyte trafficking across endothelium [11] . The ability of an individual cell to interact with and respond to ligands appears to depend on the array of expressed proteoglycans and the pattern of modifications of the heparan sulfate chains [12]–[14] . Proteoglycan synthesis in mammals is a complex process involving a diverse array of gene products including core proteins , enzymes that initiate and elongate the polysaccharide chain , and enzymes that modify the polymer by sulfation and other processes [15] , [16] . Drosophila melanogaster contains a repertoire of GAG structures similar to that of mammals [17] through a relatively small number of genes highly homologous to those found in mammals [18] , [19] . A number of viral and bacterial pathogens functionally interact with cell surface GAGs [14] . Examples include the initial attachment phase of herpes simplex virus mediated through heparan sulfate [20] , [21] , Mycobacterium tuberculosis lung epithelial invasion through expression of a heparin-binding protein [22] , and heparan sulfate-dependent attachment and entry of intestinal epithelial cells by Listeria monocytogenes [23] . We hypothesized that direct interactions with GAGs expressed on BMECs could influence the propensity of bloodborne bacteria to breach the BBB and produce CNS infection . In the specific case of GBS , a candidate ligand for this process exists in the surface-anchored Alpha C protein ( ACP ) , which has the capacity to bind GAGs and promote bacterial entry into cervical epithelial cells in vitro [24]–[26] . Drosophila can serve as a useful model organism for analysis of mechanisms of bacterial pathogenesis and resistance [27]–[30] . We have recently shown that GBS infection can be established in Drosophila , and that overall mortality and bacterial loads are reduced in fly strains with diminished GAG expression [31] . In the present study , we use bacterial , Drosophila , and mouse mutants to achieve the aim of examining the role of bacterial-GAG interactions as mediators of BBB translocation and CNS infection . The study highlights an emerging recognition that specialized surface glial cells regulate the flow of substances into and out of the fly brain and present a functional equivalent of the BBB in Drosophila [32]–[35] . Our findings indicate that the specific heparan sulfate-binding properties of ACP promote BBB interactions and contribute to the establishment of GBS meningitis .
We previously demonstrated that pricking the Drosophila thorax with a needle that had been dipped into a concentrated slurry of GBS leads to fly death [31] . To assess the nature of bacterial dissemination from the prick site , we performed histologic examination of GBS-infected Drosophila . As shown ( Figure S1 A and B ) , wild-type ( WT ) GBS A909 spreads systemically after localized injection in the thorax . H&E stain reveals bacterial cocci in multiple sites , including muscle , fat , and the lining of the brain . These data indicate that tissue architecture is largely preserved despite widespread GBS dissemination during infection . The mutant GBS strain A909/R185A harbors a single amino acid change in ACP that significantly reduces GAG binding without affecting overall ACP structure , bacterial growth rate , or surface polysaccharide capsule expression [26] , [31] . Following pinprick inoculation of an identical dose of the A909/R185A mutant into the fly thorax , bacteria disseminate less broadly from the local site of infection , and fewer cocci are visualized at the brain lining ( Figure S1 C and D ) To specifically address whether ACP-GAG binding promotes GBS dissemination into fly heads , we compared the bacterial burden ( colony forming units , cfu ) in the heads and bodies of WT flies after infection with WT GBS A909 and A909/R185A mutant . Flies infected with mutant bacteria had a significantly lower ratio of head cfu to total ( head+body ) cfu than those infected with the WT strain ( Figure 1A; Supplemental Table S1 ) , indicating that disruption of ACP-GAG binding decreases GBS penetration into the fly head . The Drosophila genome contains genes encoding the HSPG core proteins dally , dally-like protein [dlp] and syndecan [sdc]; three genes encoding heparan sulfate polymerases , tout-velu [ttv] ( homolog of mammalian Ext1 ) , sister of ttv [sotv] ( homolog of Ext2 ) and brother of ttv [botv] ( homolog of Extl3 ) ; and one gene encoding an N-deacetylase-N-sulfotransferase ( sulfateless [sfl] , a homolog of Ndst1 ) . After chain initiation by Botv , polymerization occurs through a copolymerase complex consisting of Ttv and Sotv [36] . The GAG chains then undergo a series of modifications such as N-deacetylation and N-sulfation of N-acetylglucosamine residues by Sfl . Mutations in these genes lead to altered content or sulfation of heparan sulfate on all Drosophila heparan sulfate proteoglycans [37] , [38] . To corroborate the role of GAG binding in GBS dissemination into the fly head , we used the pinprick method to establish infection in three different HSPG mutant fly strains deficient in membrane HSPG core proteins ( dally+dlp+sdc ) , heparan sulfate polymerases ( ttv+sotv ) , or NDST ( sfl ) , respectively . After infection with WT GBS A909 , all three HSPG mutant fly strains exhibited lower head/total body cfu ratios than did the WT control flies ( Figure 1B; Supplemental Table S1 ) . The extent of reduction in head/total body cfu probably underestimates the importance of HSPGs in this model since each of the host mutations had to be tested as heterozygotes due to the requirement for HSPGs in fly embryogenesis . We next sought to specifically link the observed GAG-dependent phenotype to glial cells that form tight junctions and restrict paracellular diffusion between the fly circulatory system and brain , thereby representing the Drosphila BBB equivalent [33] . The glial cell-specific driver repo-Gal4 was used to suppress polymerase expression in fly BBB cells through in vivo shRNA knockdown of ttv , sotv , and botv . After infection with WT GBS A909 , flies with glial cell-specific knockdown of either botv alone , or both ttv and sotv displayed higher survival rates than the corresponding repo and Gal4-UAS controls ( Figure 1 C and D ) . Mice heterozygous for Ext2 provide a functional in vivo model in which the chain length of GAGs expressed by the host is significantly reduced [44] . To confirm the utility of this model for analysis of GBS-BBB interactions , we isolated and cultured primary murine BMECs ( mBMECs ) from Ext2+/− mice ( Ext2 hets ) and WT littermate controls and infected them ex vivo with WT GBS A909 . We documented an approximate 40% reduction in heparan sulfate content of endothelium isolated from these Ext2 het mice vs . WT controls ( Figure S3 ) . GBS adherence and invasion were significantly reduced in mBMECs from Ext2 hets compared to those isolated from WT mice ( Figure 3 A and B ) . Several factors may explain why adherence was altered in Ext2 het mBMECs but not in Ext2-silenced hBMECs , including differences in residual GAG quantity or structure or overall surface charge , other effects of using transformed vs . primary cells , or species-specific effects . Of note , Ext2-silencing was associated with reduced hBMEC growth rate , while Ext2 het mBMECs grew similarly to WT mBMECs . Moreover , impaired GBS adherence and invasion was also observed in Ndst1-deficient mBMECs that exhibit overall reduction in sulfation of the chains ( Figure 3 C and D ) [36] , [37] . Comparable results were obtained in Ndst1-deficient mouse lung endothelial cells ( Figure S2 C and D ) . To explore a potential contribution of mammalian GAG expression to GBS BBB penetration in vivo , Ext2 hets and WT littermate controls were infected intravenously with WT GBS A909 and sacrificed 24 h later . Significantly fewer bacterial cfu ( 9-fold decrease ) were recovered from the brains of GBS-infected Ext2 hets compared to brains of WT controls ( Figure 3E ) and some animals showed dramatic 103–104 fold reduction in brain cfu . In contrast , lesser decreases in bacterial cfu were observed in the blood ( Figure 3F ) and spleen ( Figure 3G ) of Ext2 hets . Consequently , the mean ratio of brain∶blood cfu in Ext2 hets ( 10 . 6 ) was significantly lower than that observed in WT controls ( 116 . 0 ) ( Figure 3H ) . In a small pilot experiment , we noted a trend of reduced CFU in the lungs of Ext2 hets compared to WT mice ( Figure S4 ) , suggested that the finding may extend to other endothelium , but such conclusions require future corroboration . Bone-marrow derived macrophages from Ext2 hets and WT controls did not differ in assays of phagocytosis , total bacterial killing , intracellular bacterial killing , or release of tumor necrosis factor-alpha when challenged with GBS ex vivo ( Figure S5 ) , indicating that reduced GAG expression does not globally compromise phagocyte innate immune function . In sum , the reduction of GAG expression resulting from EXT2 heterozygosity diminishes GBS interactions with BMECs in vitro and decreases GBS CNS entry in vivo .
Our combined analyses in the Drosophila and mammalian systems indicate that GBS interaction with GAGs promotes BBB attachment and bacterial entry into the CNS . The GAG binding property of the surface-anchored ACP is one contributor to this invasive phenotype . Meningitis is a dangerous complication of neonatal GBS infection and understanding the predilection of the pathogen to breach the BBB is a critical goal of molecular pathogenesis investigations . To date , a number GBS factors have been shown to promote its interaction with human BMECs in vitro , including fibrinogen adhesin FbsA [38] , laminin-binding protein Lmb [45] , major pilus backbone subunit PilB [46] , lipoteichoic acid anchoring enzyme IagA [47] , and the serine-rich repeat 1 glycoprotein Srr1 [48] . In the case of IagA , Srr1 , and the hBMEC-disrupting GBS ß-hemolysin/cytolysin , a contribution to meningitis in the murine model has further been demonstrated [47]–[49] . However , the present study is the first to genetically manipulate the host to identify the corresponding host receptor molecules ( GAGs ) that the GBS virulence factor ( ACP ) exploits to adhere to and invade the BBB endothelium . Mammalian heparin sulfate chain polymerization reactions are carried out by the exostosin proteins EXT1 and EXT2 [50] , [51] and subsequent modification of the chains by the bifunctional NDSTs . We found decreased GBS invasion in both hBMECs with shRNA-reduced EXT2 levels ( Figure 2E ) and primary Ext2 heterozygous mBMECs ( Figure 3B ) . In additional studies , we found reduced GBS adherence to and invasion of Ndst1-deficient mBMECs ( Figure 3 C and D ) and lung endothelial cells ( Figure S2 C and D ) . Adherence and invasion of the A909/R185A mutant to A549 human lung epithelial cells were also markedly attenuated compared to the WT GBS A909 ( Figure S2 A and B ) . Thus the GAG-ACP interaction may contribute to GBS penetration of a broader spectrum of host cell barriers , and GAG expression patterns may determine the nature and efficiency of bacterial dissemination during infection . Moreover , our findings suggest that the chain lengths and negative charges on the GAG chains are both important to provide the binding forces between ACP and GAGs . These findings are consistent with prior data demonstrating that specific positively charged residues of ACP are required for full ACP-GAG binding affinity [26] . Whether ACP binds to a specific sequence of sulfated sugars within heparan sulfate remains to be determined . The availability of mutants altered in 2-O-sulfation of uronic acids , and 3-O- and 6-O-sulfation of glucosamine residues will allow further studies on the structural specificity of the interaction in flies [52] , [53] and in mice [54] , [55] . Our parallel findings in flies and in mammals both in vitro and in vivo support the relevance of a Drosophila infection model for the study of human CNS infections . Others have reported that the Drosophila humoral/CNS barrier conserves the xenobiotic exclusion properties of vertebrate vascular endothelium . Specifically , the exclusion process is mediated in part by a fly ATP binding cassette ( ABC ) transporter , Mdr65 , that functions similarly to mammalian xenobiotic blood-brain barrier transporters . Thus , CNS chemoprotection involves both conserved molecular structures and functionally analogous anatomic spaces that together promote CNS selective drug partition [34] . Our data extend these findings by demonstrating that CNS penetration of microorganisms may also occur via conserved molecular structures ( GAGs ) , and that these structures can be studied effectively in a Drosophila infection model . Many microorganisms display GAG-binding ability , including the two leading bacterial pathogens associated with meningitis in older children and adults: Streptococcus pneumoniae utilizes heparin , heparan sulfate and chondroitin/dermatan sulfate in colonization of respiratory mucosal epithelial cells [56] , and Neisseria meningitidis surface protein OpC binds HS proteoglycans to initiate epithelial cell invasion [57] . Sulfated GAGs also promote adherence of the veterinary meningeal pathogen Haemophilus somnus to bovine BMEC in vitro [58] . However , each of these GAG-binding interactions has uncertain functions in the pathogenesis of CNS infection that could be difficult to pinpoint because of functional redundancy of bacterial adhesins/invasins and the challenges of manipulating GAG structure/expression in complex mammalian host systems . The importance of positively charged residues is a common theme among GAG-binding proteins . For example , the C-terminal regions of mycobacterial heparin-binding hemagglutinin [59] and histone-like Hlp protein [60] contain Arg/Lys-rich repeats important for heparin binding . In some instances , common heparin-binding motifs , such as BBXB , BBBXXB ( B representing a basic amino acid residue ) , and a 20 Å spacing of basic residues have been reported [61] , [62] . Positively charged residues of ACP that were confirmed to contribute to GAG binding by site-directed mutagenesis were R172 , R185 and K196; two of these ( R185 and K196 ) are completely conserved in other members of the alpha-like protein ( Alp ) including Rib , Alp1 , Alp2 , Alp3 and Alp4 of GBS and R28 of group A Streptococcus [26] . The Drosophila infection model offers the advantages of a simpler genome and well-developed molecular approaches that will facilitate interrogation of host-pathogen GAG-binding interactions , and allow testing of candidate inhibitors of these interactions with an eye toward future therapeutic applications . The design of inhibitors targeting particular pathogen virulence mechanisms represents an attractive strategy in the era of increasing resistance to conventional antibiotics . In attempting to block bacterial-GAG interactions , a major limitation to competitive inhibition by heparin itself is its potency as an anticoagulant and the risk of hemorrhagic complications . However , synthetic low-molecular weight heparins or analogs devoid of anticoagulant activity could be contemplated in this context as potential adjunctive agents for infectious disease therapeutics .
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 California , San Diego ( Animal Welfare Assurance Number: A3033-01 ) . All efforts were made to minimize suffering of animals employed in this study . The ACP-expressing human serotype Ia GBS neonatal isolate , A909 , was used in this study . A909/R185A is an A909 point mutant strain that has an ACP variant with diminished GAG binding affinity [26] . GBS were grown in Todd-Hewitt broth ( THB , Difco ) at 37°C . SV40 large T antigen immortalized human brain endothelial cell line ( hBMEC ) was obtained from Kwang Sik Kim ( Johns Hopkins University , Baltimore , MD ) . hBMECs were maintained in RPMI 1640 medium ( Invitrogen ) supplemented with 10% FBS , 10% NuSerum ( BD ) , and 1% MEM nonessential amino acids , and were incubated at 37°C in 5% CO2 . Murine brain microvascular endothelial cells ( mBMEC ) were isolated from cerebral cortex as described [63] , except that cells were selected with 5 µg/ml puromycin for 4 days in low-glucose DMEM medium supplemented with 20% FBS ( Atlanta Biologicals ) , 50 µg/ml endothelial growth supplement ( BTI ) , 50 µg/ml heparin , nonessential amino acids , penicillin and streptomycin . Cell purity was higher than 98% as assessed by blood endothelial markers including CD31 , CD34 , CD105 and CD166 . Primary cells were cultured for 5 days and passaged once for experiments . Lung microvascular endothelial cells were isolated as described previously [64] . Ext2+/− mice were described previously [44] . Endothelial cells lacking Ndst1 were derived from Ndst1f/fTie2Cre+ mice [64] , [65] . In infection experiments , bacteria were grown to an optical density at 650 nm of 0 . 3 , and then concentrated 10-fold to approximately 2×109 cfu/ml; the exact bacterial concentration was confirmed in each experiment . Adult male flies ( 2 to 5 days old ) were anesthetized with CO2 and then pricked in the dorsal thorax underneath the wing with a fine needle previously dipped in THB broth or a concentrated solution of GBS in THB . After infection , flies were incubated at 29°C in vials with food , and fly survival was monitored over the following 4 days . Cumulative survival curves were derived , and the median survival time for each group was determined using Kaplan-Meier survival analysis . A log rank test was performed to compare survival curves . To determine the bacterial load in fly heads and bodies at 24 h after infection , flies were placed on ice , and fly heads were separated from fly bodies by a sterile surgical blade . The heads and bodies of 10 flies per group were homogenized in 500 µl and 1000 µl of phosphate-buffered saline ( PBS ) with 0 . 025% Triton X-100 respectively . The homogenates were diluted in series ( usually 10−1 to 10−3 ) , and the dilutions were plated on THB plates and incubated overnight at 37°C for cfu counting . BMECs were split into 24-well plates and allowed to grow to confluence for 48 h prior to assays , to ensure similar cell numbers for each experiment . Confluent monolayers were incubated with log-phase grown bacteria at an MOI of 1 or 10 , and centrifuged at 1600 rpm for 5 min to initiate contact . After 2 h incubation , the monolayers were washed , and 1 mL of media containing 100 µg of gentamicin and 5 µg of penicillin G was added for an additional 2 h . After washing , monolayers were disrupted by 0 . 025% Triton X-100 , and the number of invasive bacteria was quantified by serial dilution plating . To assess the level of surface-adherent ( total cell-associated ) bacteria , bacteria were quantified after 30 min of incubation without addition of antibiotics . All cellular adherence and invasion assays were performed in triplicate and repeated at least 2–3 times . Human ext2 lentiviral shRNA construct ( TRCN0000039849 ) was purchased from Open Biosystems . The resulting viruses were produced by cotransfection of 293T cells with the shRNA plasmid and packaging vectors ( Open Biosystems ) according to the vendor's instruction . The knockdown efficiency was determined by qRT-PCR analysis of ext2 expression . Primers used for qRT-PCR were EXT2 forward , 5′-AAGCACCAGGTCTTCGATTACC-3′ and reverse , 5′-GAAGTACGCTTCCCAGAACCA-3′ . and GAPDH forward 5′-GAAGGTGAAGGTCGGAGTCAACG-3′ and reverse 5′-TCCTGGAGGATGGTGATGGAAT-3′ . Ext2 heterozygous and littermate controls ( 10–12 weeks ) were injected via the tail vein with 108 cfu A909 . Twenty-four h after injection , samples of blood , brain/meninges , and spleen were collected aseptically from mice after euthanasia . Bacterial counts in blood and tissue homogenates were determined by plating serial dilutions . Bacterial counts in brain and spleen samples were corrected for differences in organ weight . Brain bacterial counts were corrected for blood contamination using the blood concentration and a conservative estimate of the mouse cerebral blood volume [47] . In a pilot experiment , additional Ext2 heterozygous and littermate controls were injected intravenously with 1×108 cfu of WT GBS and lungs harvested at 24 h for cfu determination . The significance of differences between treatment groups was determined using the unpaired Student t test .
|
Streptococcus agalactiae ( Group B Streptococcus , GBS ) is a leading cause of meningitis in human newborn infants . The bacterial and host factors that allow this pathogen to cross the blood-brain barrier ( BBB ) and cause central nervous system ( CNS ) infection are not well understood . Here we demonstrate that GBS expresses a specific protein on its surface that can bind to sugar molecules known as glycosaminoglycans ( GAGs ) on the surface of brain capillary cells , initiating infection of the BBB . Fruit flies or mice genetically engineered to have reduced GAGs showed decreased dissemination of GBS into the brain tissues following experimental infection . Our results identify a role for bacterial-GAG interactions in the pathogenesis of newborn meningitis and highlight how the simpler yet genetically conserved fruit fly GAG biosynthetic pathways make the fruit fly a good model organism to screen candidate molecules that can interrupt pathogen-GAG interactions for future therapeutic applications .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"streptococci",
"microbiology",
"host-pathogen",
"interaction",
"animal",
"models",
"bacterial",
"diseases",
"drosophila",
"melanogaster",
"model",
"organisms",
"bacterial",
"pathogens",
"infectious",
"diseases",
"infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"biology",
"mouse",
"streptococcal",
"infections",
"gram",
"positive",
"meningitis"
] |
2011
|
Glycosaminoglycan Binding Facilitates Entry of a Bacterial Pathogen into Central Nervous Systems
|
Tight regulation of immune responses is not only critical for preventing autoimmune diseases but also for preventing immunopathological damage during infections in which overactive immune responses may be more harmful for the host than the pathogen itself . Regulatory T cells ( Tregs ) play a critical role in this regulation , which was discovered using the Friend retrovirus ( FV ) mouse model . Subsequent FV studies revealed basic biological information about Tregs , including their suppressive activity on effector cells as well as the molecular mechanisms of virus-induced Treg expansion . Treg suppression not only limits immunopathology but also prevents complete elimination of pathogens contributing to chronic infections . Therefore , Tregs play a complex role in the pathogenesis of persistent retroviral infections . New therapeutic concepts to reactivate effector T-cell responses in chronic viral infections by manipulating Tregs also came from work with the FV model . This knowledge initiated many studies to characterize the role of Tregs in HIV pathogenesis in humans , where a complex picture is emerging . On one hand , Tregs suppress HIV-specific effector T-cell responses and are themselves targets of infection , but on the other hand , Tregs suppress HIV-induced immune hyperactivation and thus slow the infection of conventional CD4+ T cells and limit immunopathology . In this review , the basic findings from the FV mouse model are put into perspective with clinical and basic research from HIV studies . In addition , the few Treg studies performed in the simian immunodeficiency virus ( SIV ) monkey model will also be discussed . The review provides a comprehensive picture of the diverse role of Tregs in different retroviral infections and possible therapeutic approaches to treat retroviral chronicity and pathogenesis by manipulating Treg responses .
Seminal experiments in 1995 proved the existence of a subset of T cells termed regulatory T cells ( Tregs ) , with immunosuppressive properties critical for the control of autoimmune diseases [1] . Tregs have been demonstrated to suppress both the proliferation and function of effector T-cell subsets . They express the forkhead box protein 3 ( Foxp3 ) transcriptional factor , which is the master regulator of the suppressive program ( reviewed in [2] ) . In addition , Tregs generally express CD25 , the high-affinity receptor for interleukin 2 ( IL-2 ) , which is essential for their development and maintenance [3–5] . Tregs have been subdivided into many subsets , but we will primarily discuss the two main subpopulations of Tregs , thymic Tregs ( tTregs; previously called natural Tregs ) [6] , and peripherally derived Tregs ( pTregs; previously called induced Tregs ) . tTregs arise as Foxp3+ Tregs directly from the thymus , are generally specific for self-antigens , require continuous antigenic stimulation for survival , and act to preserve self-tolerance [1 , 7–9] . pTregs are converted to Foxp3-expressing Tregs from conventional CD4+ T cells in the periphery [10 , 11] and thus are likely to be specific for a foreign antigen . In addition to suppression of autoimmune reactivity , Tregs have also been shown to play an important role in immune evasion by cancer cells [12–14] . Therefore , the removal or blockage of Tregs is currently under investigation as a tumor therapy [14] . In 2001 , experiments in mice infected with the mouse retrovirus Friend virus ( FV ) demonstrated for the first time that Tregs were also involved in infectious diseases [15] , a finding that seemed paradoxical at the time . Subsequent studies demonstrated that Tregs were part of the normal immune response to pathogenic challenges with a number of various pathogens , including viruses , bacteria , and parasites ( reviewed in [11 , 16–18] ) . Such Treg responses are essential control mechanisms that appear to have evolved to prevent pathological damage from overly exuberant immune responses . The immunosuppressive activity of Tregs during infections both slows and dampens adaptive immune responses . For example , depletion of Tregs during acute FV infection doubles the number of virus-specific CD8+ T cells at the peak of infection and reduces viral loads by more than 10-fold [19] . Therefore , there is a trade-off between rapid and complete control of infection on one hand and minimizing inflammatory tissue damage on the other . An adverse consequence of Treg activity , especially suppression of the CD8+ T cell response , is the establishment and maintenance of chronic infection , as demonstrated in the FV model and suggested in HIV infection . Kinetic studies in the FV model indicated that Tregs become activated and significantly expanded between one and two weeks post-infection ( wpi ) [20] . Interestingly , the expansion of CD4+ Tregs during FV infection is compartmentalized in tissues with high viral replication [21] . In those tissues ( spleen , lymph nodes [LN] , and blood ) , activated Tregs remain at high frequencies throughout the course of chronic FV infection , correlating with the presence of dysfunctional virus-specific CD8+ T cells [22] . In contrast , mouse livers contain relatively few Tregs with significantly greater proportions of functional CD8+ T cells and 10-fold less chronic infection [22] . The effect of HIV infection on the frequency of Tregs has been extensively studied . Human Tregs are usually defined as CD3+/CD4+/CD25hi/CD127lo/Foxp3+ T cells . In chronic , progressive HIV or simian immunodeficiency virus ( SIV ) infections , CD4+ Tregs are more frequent in the LN and gut-associated lymphoid tissue ( GALT ) , where these viruses replicate most efficiently . Treg frequency correlates with viral loads and disease progression in HIV-infected individuals [23–26] . Treg increases occur early during the course of infection , as shown in the SIV monkey model [27] . Moreover , Tregs are activated during chronic HIV infection , with higher expression of molecules associated with activation , such as CD39 or cytotoxic T lymphocyte–associated protein 4 ( CTLA-4 ) [23 , 28 , 29] . However , in contrast to FV infection , it should be noted that the overall number of Tregs decreases during chronic HIV infection , although Tregs remain selectively spared compared with other CD4 subsets [23 , 29] . Highly active antiretroviral therapy ( HAART ) only partially abolishes the effect of HIV infection on abnormal Treg frequency and phenotypic characteristics [23 , 29–31] . The effect of HIV infection on a subset of Tregs , the follicular regulatory T cells ( Tfr ) , remains uncertain . These cells were first described in 2011 and control germinal center responses [32 , 33] . Both human and murine Tfr display a unique transcriptional pattern overlapping that of both follicular T helper cells ( Tfh ) and Treg , notably coexpressing B cell lymphoma 6 protein ( Bcl-6 ) , Foxp3 , and B lymphocyte–induced maturation protein 1 ( Blimp-1 ) . These C-X-C chemokine receptor 5 positive ( CXCR5+ ) Tfr regulate the magnitude and character of the antibody ( Ab ) response by limiting the size of the Tfh compartment , inhibiting the selection of germinal center B cells , or both . Due to the high interest in the mechanisms regulating the development of broadly neutralizing Ab to HIV ( reviewed in [34] ) , the function and homeostasis of these cells during HIV and/or SIV infection have been studied by several groups , but the data reported thus far are contradictory . Indeed , the frequency of LN or splenic Tfr was described in chronically SIV-infected rhesus macaques as decreased [35] , unchanged [36] , or increased [37] and as increased in HIV-infected individuals [38] .
Because Tregs express a normal T-cell receptor ( TCR ) , it was originally thought that they might simply recognize and respond to viral antigens . That hypothesis , however , could not explain how or why the immune system would recognize one antigen as stimulatory and another as suppressive . In contrast with the theory of pathogen-specific recognition by Tregs , we were unable to demonstrate their presence in FV-infected mice despite many attempts using major histocompatibility complex ( MHC ) class II tetramers and TCR transgenic FV-specific CD4+ T cells [39 , 40] . In fact , FV-specific TCRs are specifically excluded from the Treg repertoire in mice [39 , 40] . Instead , FV-induced Tregs display a very broad distribution of TCR variable β ( Vß ) chain usage , suggesting that they recognize a wide variety of different self-antigens [39] . This is not too surprising because tTregs are generally specific for self-antigens , require continuous antigenic stimulation for survival , and act to preserve self-tolerance [1 , 7–9] . Therefore , tTregs are quite different from conventional CD4+ T cells , which circulate and survive in a naïve state without TCR stimulation . tTregs are distinct from pTregs , which are converted to Foxp3-expressing Tregs from conventional CD4+ T cells in the periphery and may be virus specific [10 , 11] . Virus-specific Tregs have been described in a few human infections , including HIV and hepatitis C virus ( HCV ) , but they appear to be so infrequent in those infections that their biological relevance is questionable [41–44] . A side-by-side comparison of the TCR repertoires of Tregs versus conventional CD4+ T cells in HIV-infected individuals has not been done , although the analysis of purified Tregs from HIV-infected patients showed an overrepresentation of some Vα and Vβ families when compared to the Treg repertoire in uninfected individuals [45] . However , such overrepresentation does not appear to be Treg specific because it was also reported in unfractionated CD4+ T cells from HIV-infected individuals [46] . Two distinct mechanisms of Treg expansion have been defined in the FV model , one IL-2-dependent ( Fig 1 ) and the other IL-2-independent ( Fig 2 ) . Tregs express high levels of the IL-2 receptor ( CD25 ) , IL-2 is an essential differentiation factor for Tregs [47] , and it is generally required for Treg function [3–5] . Therefore , it is not surprising that IL-2 is a required secondary signal involved in the expansion of most FV-induced Tregs [39] . In FV infections , IL-2 is predominantly produced by FV-specific effector CD4+ helper T cells responding to the infection [40] . It was recently shown that this IL-2-dependent Treg expansion is also dependent on interactions with B cells [48] . B cell–dependent Treg signaling occurrs via tumor necrosis factor ( TNF ) receptor superfamily member 18 ( glucocorticoid-induced TNF receptor-related protein [GITR] ) ligation with GITR ligand ( GITRL ) on B cells [48] . Of note , GITR–GITRL interactions are also required to control autoimmunity through regulation of Treg homeostasis [49] . The expanded Tregs in FV infection are tTregs as defined by their expression of the markers Foxp3 , CD25 , HELIOS , and Neuropilin 1 [39] , as well as the fact that responding Tregs arise from preexisting tTreg populations and no conversion of conventional T cells into Tregs occurs [39] . A second mechanism of Treg induction , which is self-antigen specific but IL-2-independent , accounts for about 10% of the Treg expansion during FV infection ( Fig 2 ) [39 , 50] . Analyses of TCR Vß chain usage showed that a subpopulation of Tregs expressing the Vß5 chain of the TCR expanded disproportionately after FV infection [39] . This Treg subpopulation is specific for an endogenous retroviral superantigen ( Sag ) encoded by the mouse mammary tumor virus 9 ( MMTV9 ) . MMTV9 Sag binds to all CD4+ T cells expressing Vß5 chains and delivers a potent primary TCR signal that causes deletion of conventional CD4+ T cells during thymic selection in order to prevent autoreactivity [51 , 52] . However , tolerogenic Foxp3+ Tregs are not deleted [53] , and Sag stimulation results in the up-regulation of TNF receptor 2 ( TNFR2 ) on the cell surface of Vß5+ Tregs in the periphery [50] . If TNFR2 on these Tregs binds the membrane-bound form of TNFα [50] , it provides signal 2 for Treg activation . Interestingly , the membrane-bound form of TNFα is transiently up-regulated on recently activated effector CD8+ T cells , which in the case of FV infection are FV-specific CD8+ T cells [50] . Therefore , it is the effector CD8+ T cells , which eventually become the targets of Treg-mediated suppression , that provide the second signal for the activation and proliferation of Vß5+ Tregs . Apparently , the combination of a potent Sag signal combined with TNFR2 signaling is strong enough to negate the normal requirement for IL-2 . This IL-2-independent mechanism does not appear restricted to FV infection because the Vß5 subset of Tregs also disproportionately expands in mice persistently infected with lymphocytic choriomeningitis virus ( LCMV ) [54] . Treg subpopulations in humans also express TNFR2 [55–57] , but it remains to be determined whether Tregs that are specific for endogenous retroviral antigens also exist in humans . Along this line , it is interesting that human endogenous retroviral ( HERV ) -specific conventional T cells expand in HIV-infected individuals [58] . Mechanisms promoting the expansion of Tregs during HIV infection are not clearly understood , but it appears to be more a relative sparing than a real expansion , i . e . , fewer CD4+ Tregs are killed by HIV than conventional CD4+ T cells . The field has not yet come up with a satisfactory explanation of why absolute numbers of Tregs decrease even though , based on all the mechanisms studied so far , an expansion would be expected . First , ex vivo and in vitro studies suggest that the proportion of pTregs may increase during HIV infection . One potential mechanism for this enhanced conversion is that HIV or SIV infection induces semimature dendritic cells ( both myeloid and plasmacytoid DCs ) that have been shown to enhance Treg differentiation from conventional CD4+ T cells [59 , 60] . In vitro culture with myeloid DCs from HIV-infected individuals also promotes Treg expansion [61] . Virus-infected DCs are also involved in Treg expansion in FV infection , although there is no conversion of conventional T cells into pTregs [62] . Second , Treg proliferation appears augmented during chronic HIV infection , as Tregs from chronically infected patients express higher levels of Ki-67 than those from uninfected individuals [29 , 63] . Finally , Tregs also seem less prone to HIV-induced apoptosis [26 , 64 , 65] , which is consistent with the fact that Tregs express lower levels of proapoptotic molecules than their conventional CD4+ T-cell counterparts in the gut of SIV-infected rhesus macaques [66] . However , all these ideas are based on ex vivo or in vitro studies because the tools to ascertain whether these pathways are operational in vivo are lacking in humans and nonhuman primates .
CD8+ T cells are extremely potent effector cells that not only secrete potent inflammatory cytokines but also kill infected cells through the release of cytotoxic granules , including perforin and granzymes . Therefore , they have the potential to cause significant collateral damage during a host immune response and were the first antiviral cells recognized as targets for Treg-mediated suppression [15 , 67] . Studies using adoptively transferred TCR-transgenic , FV-specific CD8+ T cells [67] , as well as studies in which mice could be selectively depleted of Tregs [19 , 21] , indicated that Tregs begin to suppress CD8+ T-cell proliferation and effector functions during the late phase of acute FV infection [21] . Tregs mainly affect the exocytosis pathway of CD8+ T-cell killing rather than the first apoptosis signal receptor ( Fas; or CD95 ) /Fas ligand ( FasL ) pathway [68] . Treg-mediated suppression is maintained during chronic FV infection and contributes to the exhausted phenotype of CD8+ T cells [69] . Therefore , selective depletion of Tregs during chronic FV infection reactivates residual FV-specific CD8+ T cells to secrete multiple cytokines , produce cytotoxic granules , and develop in vivo cytotoxicity resulting in significantly reduced chronic viral set points [69] . Interestingly , the chronic exhaustion of FV-specific CD8+ T cells is also influenced by the expression of inhibitory receptors [70 , 71] , a separate immune checkpoint mechanism that can act independently of Treg responses [70 , 71] . Importantly , the suppressive activity of FV-induced Tregs on CD8+ T cells is not antigen specific . After becoming activated and expanded during FV infection , Tregs can suppress ovalbumin-specific CD8+ T cells or mixed lymphocyte reactions in vitro [15 , 72] , and they also impair mouse CMV-specific T-cell responses in vivo [73] . There is substantial experimental evidence that both Tregs and inhibitory receptor expression play key roles in T-cell exhaustion and immune dysfunction during chronic HIV and SIV infections [70 , 74–76] . Early in vitro studies using Tregs from HIV-1+ patient samples showed that both HIV- and CMV-specific CD8+ T-cell responses were suppressed [77–79] . Importantly , in vivo studies have also illustrated Treg effects on CD8+ T cells . For example , studies done by the Apetrei-Pandrea group used the human IL-2/diphtheria toxin fusion protein ( Ontak ) to deplete Tregs in SIV-infected controller macaques [80] . Following this treatment ( leading to a >75% loss in Treg proportion ) , major CD4+ T-cell activation occurred , leading to the reactivation of latent SIV . However , Treg depletion also significantly boosted SIV-specific CD8+ T-cell frequencies , resulting in the rapid clearance of reactivated virus . These data demonstrate the complex in vivo role of Tregs in controlling SIV-specific immune responses . These results also support the concept that the early emergence and persistent accrual of Tregs in LN during pathogenic SIV and/or HIV infection likely impairs the protective antiviral CD8+ T-cell response , as suggested by previous associative studies ( Fig 3 ) [26 , 27 , 81] . Also of interest in this context is the fact that CD8+ T cells restricted by the human leukocyte antigen ( HLA ) allele groups associated with delayed HIV disease progression ( notably HLA-B*27 and HLA-B*57 ) were not suppressed ex vivo by Tregs [82] . This resistance to suppression implies that Treg accrual plays a role in HIV-1 disease progression by hampering CD8+ T-cell responses . As described in FV infection , Treg functions are not virus specific in HIV infection because they also control CD8+ T-cell responses against other viruses , such as CMV [31 , 83] . Therefore , their increased frequency likely contributes to the development of AIDS-associated infections in untreated patients . Accordingly , a higher percentage of circulating CD4+/FOXP3+ Tregs was shown to be predictive of CMV end-organ disease [84] . Effector CD4+ T cells , responsible for providing help for effector B and T-cell responses and the development of immunological memory , are also targeted by Tregs in the FV model . Tregs can suppress the proliferation and cytokine production by type I helper T cells during FV infection [85] . In addition , they also control the cytotoxicity of virus-specific CD4+ T cells during acute FV infection but only in circumstances in which CD8+ T cells are absent . If cytotoxic CD8+ T cells are present , Treg depletion does not result in a substantial induction of CD4+ T-cell cytotoxicity [86] . However , in acutely FV-infected mice lacking both CD8+ T cells and Tregs , a massive expansion of cytotoxic CD4+ T cells occurs . These CD4+ T cells can kill FV antigen-labeled targets in an MHC class II–dependent exocytosis process . Such CD4+ T cell–mediated cytotoxicity appears to be a mechanism to compensate for the lack of CD8 functionality in chronically FV-infected mice [87] . In chronic FV infections in which CD8+ T cells are dysfunctional , cytotoxic CD4+ T cells take over and kill virus-infected targets in a Fas/FasL-mediated pathway [88] . During HIV/SIV infection , Tregs decrease CD4 functionality , inhibiting both HIV-specific and polyclonal responses ( reviewed in [89–91] ) . However , the role of Tregs is complex ( Fig 3 ) because low Treg frequency during HIV infection is associated with increased immune activation [92–94] , and such activation facilitates HIV infection of target cells . Therefore , Tregs also limit virus spread . This has been shown by in vivo Treg depletion of SIV-infected controller macaques as mentioned above [80] and also in the nonpathogenic model of SIV infection in African Green Monkeys [95] . It has also been shown in vitro that Tregs limit the HIV infection of activated conventional CD4+ T cells [96] , macrophages [97] , and DCs [98] . Tfr , which suppress Tfh necessary for B cell help in germinal centers [99] and subsequent Ab responses [100 , 101] , have been the subject of several recent studies [34 , 102 , 103] . Tfr normally differentiate from tTregs rather than conventional antigen-specific T cells [104] , although exceptions have been observed [105] . The frequency of Tfr in LN is inversely correlated with the frequency of Tfh and germinal center B cells in LN [35 , 36] , and also with the avidity of plasma Abs recognizing SIV envelope proteins [35] . Ex vivo HIV infection of human Tfr increased their expression of molecules associated with regulatory activity , namely CTLA-4 , lymphocycte activation gene 3 ( LAG-3 ) , GITR , and galectin-3 , and also enhanced production of IL-10 and transforming gowth factor β ( TGF-β ) . Consequently , these infected Tfr are very efficient at inhibiting Tfh proliferation and Tfh production of cytokines for B cell help [37] . Similarly , transcriptional analysis of LN Tfr after SIV infection revealed a profile of increased activation [36] . In the same line , HIV-1-infected individuals with broadly neutralizing Abs had a lower frequency of Tregs and a higher frequency of circulating memory Tfh compared with those who did not develop such Abs [106] . This increased response was not specific for HIV antigens because the individuals with broadly neutralizing Abs also had a higher frequency of auto-Abs [106] . Natural killer ( NK ) cells are important innate lymphoctyes that produce proinflammatory cytokines and can kill tumor cells and infected cells following their activation . It has been shown that Tregs indirectly regulate NK cell maturation by restraining IL-2 availability [107] . Therefore , Tregs can regulate the ability of NK cells to react to missing self-antigens on target cells , essentially acting as an IL-2 sink [108] . In a related manner , Tregs regulate NK responses during FV infection . Normally , NK cell responses to acute FV infection are rather weak , and NK cells contribute only marginally to virus control [109] . However , a large part of this impotence is because FV-induced Tregs suppress NK cell proliferation , maturation , and effector cell differentiation during the acute phase of FV infection [110] . Because Tregs express high levels of the high-affinity IL-2 receptor CD25 while NK cells only express the low-affinity IL-2 receptor CD122 , Tregs can outcompete NK cells for IL-2 consumption . This lack of IL-2 availability reduces the activation and differentiation of NK cells in FV-infected mice . When Tregs are depleted or IL-2 is experimentally directed to the CD122 receptor of NK cells , full activation of NK cells and significant anti-FV activity are observed . These results indicate that targeted immunotherapy can abrogate the suppression of NK cells by Tregs and enhance virus control . To date , nothing has been published about the influence of Tregs on NK cell responses in HIV-infected humans . Tregs also mediate their suppressive action by acting directly on antigen-presenting cells , such as DCs , decreasing DC maturation and subsequently T-cell activation [111] . The formation of Treg–DC conjugates in vivo and in vitro also suggests that DCs may be primary targets of Treg suppression [111–115] . Because DCs facilitate HIV dissemination to the lymphoid organs by enabling HIV infection of CD4+ T cells [116 , 117] , we analyzed how Tregs affect viral transmission from DC to effector T cells . In an in vitro model , Tregs significantly decreased HIV transmission from DC to T cells , notably impairing actin polymerization and the trafficking of HIV viral particles to the immunological synapse [98] . Using mice infected with the immunodeficiency-inducing retroviral complex known as LP-BM5 , the Green lab demonstrated an increased proportion of both IL-10-producing Tregs and immunosuppressive myeloid-derived suppressor cells ( MDSCs ) [118] . In adoptive transfer experiments , it was further shown that these two different immunosuppressive cell subsets reciprocally modulated each other’s function . MDSCs modulated IL-10 production by Tregs , and in the absence of Tregs , MDSC suppression of T cells was increased .
Because Tregs are themselves susceptible to HIV infection ( Fig 3 ) [119] , one question that has been addressed by several groups is whether their functionality , on a per cell basis , was altered by HIV infection . Of note , bulk Tregs isolated from chronic HIV-1 progressors are as functional as those from uninfected individuals or HIV controllers [94 , 120 , 121] . However , binding of inactivated HIV or HIV glycoprotein 120 ( gp120 ) to CD4 on Tregs enhanced not only their survival but also their suppressor activities [26 , 64 , 65] . In contrast , one study reported that productively infected Tregs were less functional on a per cell basis than uninfected , but HIV-exposed , Tregs [122] . Therefore , there may be a different effect from productive infection compared to exposure to defective HIV particles circulating during HIV infection .
Knowledge about the molecular mechanisms underlying Treg-mediated suppression of retrovirus-specific T-cell responses is limited . In FV infection , suppression occurs in a direct cell-to-cell contact–dependent manner independently of the presence of antigen-presenting cells [72] . Immunosuppressive cytokines such as IL-10 and TGF-β secreted by CD4+ Tregs did not contribute to Treg-mediated immunosuppression in either in vitro or in vivo experiments [67 , 72] . Furthermore , Tregs that respond to FV infection do not secrete granzymes , excluding granzyme-mediated killing of effector T cells [123] . This is in line with findings that Tregs inhibit effector T-cell proliferation and function [21] but do not induce apoptosis in T cells during FV infection . The exact mechanism of suppression by Tregs in FV-infected mice is still under investigation . Tregs have been reported to control T-cell activation and proliferation via a contact-dependent mechanism involving cyclic andenosine monophosphate ( cAMP ) [124] , a mechanism that might also be operative during FV infection . During HIV infection , CTLA-4 ( CD152 ) , CD39 , and cAMP have been intensively studied as mechanisms of suppression . CTLA-4 coinhibitory molecules are expressed by Tregs from HIV-infected individuals at higher levels than in uninfected individuals [23] . CTLA-4 blockade early during SIV infection of rhesus macaques led to an increase in T-cell activation and viral replication [125] . However , interpretation of these data must be cautious because the Treg role was not specifically addressed in this study . Moreover , Kaufmann et al . showed that Tregs did not play a major part in the in vitro CTLA-4-mediated inhibition of HIV-specific responses because CTLA-4 blockade was still operative in the absence of Tregs and most HIV-specific CTLA-4+/CD4+ T cells were not Tregs [126] . The proportion of CD39+ Tregs also increases during chronic HIV infection [23 , 28 , 29] , and in vitro blockade of CD39 reverses Treg suppression of HIV-specific CD8+ T cells [28] . This suppressive mechanism may be particularly important in vivo because both CD4+ and CD8+ T cells from chronically HIV-infected individuals express high levels of the adenosine A2A receptor CD39 [28 , 127] . Importantly , a CD39 gene polymorphism leading to low CD39 expression is associated with a slower progression to AIDS [28] . However , this association cannot be solely ascribed to CD39+ Tregs because type 1 regulatory T cells ( Tr1 ) or CD8+ regulatory cells also express CD39 , and both cell types are frequent in HIV-infected individuals [128 , 129] . As mentioned above , cAMP also participates in Treg suppression . Upon stimulation with HIV gp120 , human Tregs were shown to accumulate cAMP in their cytosol . Furthermore , the tolerizing effect of HIV gp120 in a xenogeneic graft-versus-host disease model was strictly dependent on the induction of cAMP in human Tregs [64] . In agreement with this hypothesis , Tregs from HIV-1-infected individuals express high levels of intracellular cAMP [127] . We also showed that Treg cAMP was important to control HIV replication in activated effector T cells because suppression was abolished by chemically decreased cAMP levels in Tregs . A similar effect was found in DCs [98] . Blocking gap junction formation between Tregs and effector T cells and inhibiting protein kinase A in effector T cells also abolished Treg suppression , indicating a requirement for cell-to-cell contact [96] .
Because Tregs can blunt effector immune responses during retroviral infections , it is important to determine whether Treg responses can be manipulated in vivo to overcome suppression and induce immunity . Several approaches have been studied in the FV model . As mentioned earlier , Tregs express GITR , and treatment of mice during acute FV infection using blocking anti-GITR Ab significantly increased virus-specific CD4+ and CD8+ T-cell numbers and function , reduced pathology , and produced long-term increases in CD8+ T-cell functionality [130] . A caveat to this study is that all of the effects could not be attributed to Tregs but might also have been direct activation of the CD8+ T cells . In another study , it was shown that CD8+ T cells could be rendered resistant to Treg-mediated immunosuppression by stimulating them with an agonistic Ab specific for the CD137 ( 4-1BB ) costimulatory molecule [131] . Interestingly , this CD137 agonistic Ab could also be used to reprogram Tregs to become cytotoxic CD4+ T cells with antitumor activity [132] . In those studies , the reprogrammed cells expressed the T-box transcriptional factor Eomesodermin and granzyme B without loss of Foxp3 expression . Treg responses have also been successfully manipulated to enhance the efficacy of therapeutic vaccines during chronic FV infection . Therapeutic vaccination of chronically FV-infected mice with functionalized calcium phosphate ( CaP ) nanoparticles temporarily reactivated cytotoxic CD8+ T cells and significantly reduced viral loads [133] . Transient ablation of Tregs during this nanoparticle-based vaccination strongly enhanced antiviral immunity and further decreased chronic viral set points [134] . In the context of HIV infection , Tregs also seem detrimental to the efficacy of therapeutic vaccines [135 , 136] , suggesting that adding a Treg blocker along with the vaccine might lead to higher clinical benefit although , to our knowledge , this strategy has not yet been tested . During HIV infection , another exciting prospect is that Treg manipulation could be used to purge the HIV reservoir . In the last 15 years , it has become evident that even the most highly efficient antiretroviral therapy will not cure HIV , due to the persistence of integrated , replication-competent proviruses within host cellular DNA ( rev . in [137] ) . Therefore , developing “shock and kill” strategies has emerged as a priority in the field of HIV research . The underlying concept of these strategies is that if it were possible to induce viral expression from the latent reservoir , then it might be feasible to trigger immune-mediated clearance of the infected cells through CTLs , NK cells , or immunotoxins . However , a critical limitation of these strategies is that effective and safe latency reversing agents have not yet been identified , and exhaustion of the HIV- and/or SIV-specific CTLs hampers their capacity to properly eliminate reactivated virus ( rev . in [137] ) . In this context , manipulation of Treg numbers and/or functions could be advantageous , based on the results of Treg depletion in simian models . Indeed , as mentioned above , transiently decreasing Treg numbers by Ontak treatment led to both reactivation of latent HIV and boosted SIV-specific CD8+ T-cell frequencies in virally suppressed rhesus macaques . The result was rapid clearance of the reactivated virus [80] , thus achieving both goals of the “shock and kill” approach at the same time . This treatment was safe , without signs of untoward autoimmune disease ( [80] and 2017 discussion between Drs Apetrei , Pandrea , and Chougnet ) . Transient Treg ablation might be more efficacious in this context than the manipulation of Treg function because Tregs themselves have been suggested to be a reservoir for latent HIV [138 , 139] , although that has not been confirmed by all studies [140] . Of note , in an HIV-infected patient being treated for melanoma with Ipilimumab , an anti-CTLA-4 Ab known to effect Tregs , transient decreases in HIV-1 RNA were detected following infusions of the Ab [141] . Besides HIV , Treg-mediated suppression of antiviral effector cells is a matter of concern in other chronic virus infections such as HCV , hepatitis B virus ( HBV ) , human papillomavirus ( HPV ) , and Epstein-Barr virus ( EBV ) [74] and limits responses to therapeutic vaccines in patients infected by these chronic viruses . Notably , patients with large lesions due to HPV-induced vulvar intraepithelial neoplasia also had high frequencies of HPV-specific CD4+/CD25+/Foxp3+ T cells and displayed a lower HPV-specific interferon γ ( IFNγ ) /IL-10 ratio after therapeutic vaccination [142] . Therefore , therapeutic manipulation of Tregs to reactivate or enhance virus-specific immunity and subsequently reduce chronic infection levels could have widespread clinical applications . In considering therapeutic manipulation of Tregs , the Janus nature of Tregs during HIV infection should not be forgotten . Tregs limit generalized immune activation during HIV infection [92–94] , and in the era of suppressive HAART , persistence of immune activation is highly associated with the noninfectious causes of HIV-driven increased mortality ( rev . in [143] and [144] ) . Relevant to this concept , low Treg frequency in HIV elite controllers is strongly associated not only with immune activation but also with accelerated atherosclerosis and other morbidities linked to inflammation [145 , 146] . Therefore , depletion or ablation of Treg function to boost immune responses or purge the reservoir may end up producing deleterious consequences . In this context , increasing Treg frequency could be beneficial to HIV patients because Tregs limit atherosclerosis ( rev . in [147 , 148] ) . The ongoing statin clinical trials in HAART-treated patients will be informative in this context . Statins are given to HAART-treated patients , including normolipidemic patients ( REPRIEVE Phase IV clinical trial ) , with the goal of decreasing HIV-associated cardiovascular risk . Importantly , we and others have shown that statin treatment increases Treg frequency [149 , 150] , which likely contributes to their pleiotropic anti-inflammatory properties . Analysis of Treg dynamics in statin-treated HIV-infected patients , in relation to immune activation and clinical outcome , may provide important information about the exact role of Tregs in HIV infection .
Given the complex roles that Tregs play in retroviral and other infectious diseases , the successful application of therapeutics to treat infectious diseases via modulation of Tregs will obviously require extremely detailed information regarding both the positive and negative contributions of Tregs in a particular infection . We now know that the balance of beneficial versus detrimental effects from Tregs can change during the course of a retroviral infection , especially between the acute and chronic phase of infection . Therefore , it is important to recognize not only which infectious agent induced the Treg response but also the phase of the infection . The discovery of molecular mechanisms that initiate and control Treg responses in infectious diseases is key to the understanding and manipulation of these complex processes . For example , the presence of endogenous retroviruses that express specific antigens may strongly affect individual Treg responses . We also know that virus-induced Tregs alter not only the response against the virus that initiated the response but also subsequent responses to other infections . Therefore , the patient’s history of infections can affect the homeostatic level of Treg-mediated suppression . Numerous other factors that need further study may also play important roles in Treg responses , including sex , age , underlying medical conditions , drug use , stress , etc . It is extremely important that further investigations continue to delineate the important factors influencing the impact of Tregs on both pathology and antiviral immunity in retroviral infections so that personalized medicine can be developed . Discovery of the specific mechanisms that Tregs use for immunosuppression may even allow differential blockade of detrimental functions while maintaining beneficial functions . The targeted manipulation of Treg responses holds a bright future for treating not only autoimmune diseases but also in enhancing vaccine responses , immune responses to infections , and in eradication of chronic infections .
|
Regulatory T cells ( Tregs ) play a very complex role in retroviral infections , and the balance of beneficial versus detrimental effects from Tregs can change between the acute and chronic phase of infection . Therefore , the development of therapeutics to treat chronic retroviral infections via modulation of Tregs requires detailed information regarding both the positive and negative contributions of Tregs in a particular phase of a specific infection . Here , we review the molecular mechanisms that initiate and control Treg responses in retroviral infections as well as the target cells that are functionally manipulated by Tregs . Basic findings from the Friend retrovirus mouse model that initiated this area of research are put into perspective with clinical and basic research from HIV studies . The targeted manipulation of Treg responses holds a bright future for enhancing immune responses to infections , vaccine responses , and for cure or functional cure of chronic retroviral infections .
|
[
"Abstract",
"Regulatory",
"T-cell",
"responses",
"in",
"retroviral",
"infections",
"Mechanisms",
"of",
"Treg",
"expansion",
"and/or",
"accrual",
"during",
"retroviral",
"infection",
"Targets",
"of",
"Treg",
"suppression",
"during",
"retrovirus",
"infection",
"Effect",
"of",
"HIV",
"infection",
"on",
"Treg",
"functionality",
"Mechanisms",
"of",
"suppression",
"Treg",
"immunotherapy",
"in",
"lentiviral",
"infections:",
"The",
"pros",
"and",
"cons",
"Conclusions"
] |
[
"blood",
"cells",
"hiv",
"infections",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"cytokines",
"pathogens",
"immunology",
"microbiology",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"developmental",
"biology",
"review",
"rna",
"viruses",
"molecular",
"development",
"cytotoxic",
"t",
"cells",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"hiv",
"t",
"cells",
"microbial",
"pathogens",
"immune",
"response",
"immune",
"system",
"biochemistry",
"signal",
"transduction",
"t",
"cell",
"receptors",
"cell",
"biology",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"regulatory",
"t",
"cells",
"viral",
"diseases",
"lentivirus",
"immune",
"receptors",
"organisms"
] |
2018
|
Regulatory T cells in retroviral infections
|
Forms of homeostatic plasticity stabilize neuronal outputs and promote physiologically favorable synapse function . A well-studied homeostatic system operates at the Drosophila melanogaster larval neuromuscular junction ( NMJ ) . At the NMJ , impairment of postsynaptic glutamate receptor activity is offset by a compensatory increase in presynaptic neurotransmitter release . We aim to elucidate how this process operates on a molecular level and is preserved throughout development . In this study , we identified a tyrosine kinase-driven signaling system that sustains homeostatic control of NMJ function . We identified C-terminal Src Kinase ( Csk ) as a potential regulator of synaptic homeostasis through an RNAi- and electrophysiology-based genetic screen . We found that Csk loss-of-function mutations impaired the sustained expression of homeostatic plasticity at the NMJ , without drastically altering synapse growth or baseline neurotransmission . Muscle-specific overexpression of Src Family Kinase ( SFK ) substrates that are negatively regulated by Csk also impaired NMJ homeostasis . Surprisingly , we found that transgenic Csk-YFP can support homeostatic plasticity at the NMJ when expressed either in the muscle or in the nerve . However , only muscle-expressed Csk-YFP was able to localize to NMJ structures . By immunostaining , we found that Csk mutant NMJs had dysregulated expression of the Neural Cell Adhesion Molecule homolog Fasciclin II ( FasII ) . By immunoblotting , we found that levels of a specific isoform of FasII were decreased in homeostatically challenged GluRIIA mutant animals–but markedly increased in Csk mutant animals . Additionally , we found that postsynaptic overexpression of FasII from its endogenous locus was sufficient to impair synaptic homeostasis , and genetically reducing FasII levels in Csk mutants fully restored synaptic homeostasis . Based on these data , we propose that Csk and its SFK substrates impinge upon homeostatic control of NMJ function by regulating downstream expression or localization of FasII .
Throughout a metazoan’s lifespan , its nervous system encounters numerous challenges to function . Responding to stress requires the flexibility afforded by forms of neuroplasticity . Yet even in plastic neurons , functional outputs of synapses must be kept within physiologically appropriate ranges . This type of control requires sensitive regulatory systems , including homeostatic forms of synaptic plasticity [1–7] . The precise molecular underpinnings of homeostatic neuroplasticity are elusive . Progress has been made at identifying individual molecules required for the homeostatic regulation of synapse function [1 , 6 , 7] . In addition , compelling links have been suggested between homeostatic synaptic plasticity and disease processes , such as epilepsy [8] . However , few factors have been organized into coherent signaling pathways . The Drosophila melanogaster third instar larval neuromuscular junction ( NMJ ) is an ideal synapse for the study of homeostatic regulation [9] . At the NMJ , genetic and pharmacological manipulations can be employed to decrease the sensitivity of postsynaptic receptors to single vesicles of glutamate , thereby decreasing quantal size [9–11] . Decreased quantal size triggers retrograde ( muscle-to-nerve ) , homeostatic signaling that drives increased evoked presynaptic neurotransmitter vesicle release–increased quantal content ( QC ) . As a result , relatively normal levels of muscle excitation are maintained . We refer to this compensatory process as “synaptic homeostasis” or “NMJ homeostasis . ” Acute NMJ application of the postsynaptic glutamate receptor inhibitor philanthotoxin-433 ( PhTox ) induces a rapid decrease in quantal size and a sharp homeostatic increase of presynaptic neurotransmitter release on a short timescale ( 5–10 min ) [10] . Since the retrograde and cellular signaling pathways that drive NMJ homeostatic plasticity are not well defined [9] , PhTox has been exploited in screens to identify mutations that impair the rapid induction of synaptic homeostasis . What has emerged is an exquisite array of processes that dictates increases in QC through acute alterations in presynaptic Ca2+ influx , regulation of the size and replenishment of the readily releasable pool ( RRP ) of presynaptic vesicles , control of vesicle fusion events , and intrinsic neuronal excitability [10 , 12–23] . Changes in these Drosophila NMJ parameters resemble homeostatic processes that have been documented in mammalian synaptic preparations [24–29] . Therefore , diverse homeostatic processes likely share ancient , universally conserved mechanisms . Less explored is the identity of molecules and signaling systems that sustain NMJ homeostasis throughout development . Studies indicate that some molecules required for the long-term maintenance of homeostatic plasticity at the NMJ are not required for the short-term induction of homeostatic signaling [9 , 30–33] . To address this issue in more depth , we combined RNA interference ( RNAi ) -based screening with genetics , electrophysiology , and synapse imaging [33] . Here we show that classical signaling molecules , including C-terminal Src Kinase ( Csk ) and Src family kinases ( SFKs ) , and the trans-synaptic neural cell adhesion molecule Fasciclin II ( FasII/NCAM ) have roles in the long-term maintenance of synaptic homeostasis . Our data suggest Csk promotes homeostatic plasticity , at least in part , by repressing synaptic levels of FasII/NCAM . Our data could have implications for the strategies employed by synapses to facilitate long-term , trans-synaptic homeostatic signaling throughout development .
In an ongoing RNA interference ( RNAi ) -based screen to identify factors required to sustain homeostatic plasticity [33] , we uncovered Csk as a potential regulator . For the screen , we impaired the expression of the GluRIII glutamate receptor subunit gene by RNAi , along with concurrent pre- and postsynaptic RNAi-mediated knockdown of screened target genes [33] . GluRIII knockdown on its own caused significantly diminished NMJ quantal size ( mEPSP ) ( Fig 1A ) [33] . Diminished quantal size was offset by a homeostatic increase in presynaptic neurotransmitter release ( increased quantal content , QC ) ( Fig 1A; see S1 Table for raw electrophysiological data throughout manuscript ) [33] . The response was similar to what has been shown previously for a GluRIIA glutamate receptor subunit deletion or mutation [11 , 34] , acute philanthotoxin-433 ( PhTox ) application to the NMJ [10] , and GluRIII genetic mutation [35] . By contrast , when the GluRIII and Csk genes were simultaneously knocked down by RNAi , larval progeny showed no significant increase in QC compared to Csk[RNAi] genetic controls ( Fig 1A ) . This caused severely diminished evoked muscle excitation ( Fig 1B ) . These initial findings suggested that Csk could be required for synaptic homeostasis . We turned to loss-of-function mutations in GluRIIA and Csk to confirm the RNAi results . GluRIIASP16 deletion NMJs have diminished quantal size , and the NMJ compensates for this defect with a homeostatic increase in quantal content ( Fig 1C ) [11] . We tested this canonical response for two independent , hypomorphic transposon insertions in the Csk locus , Cskc04256 and Cskj1D8 [36–39] . By electrophysiology , both GluRIIA; Cskc04256 double mutant NMJs and GluRIIA; Cskc04256/Cskj1D8 heteroallelic double mutant NMJs had expected decreases in quantal size compared to baseline Csk mutant genetic controls . However , neither had a homeostatic increase in QC ( Fig 1C and 1D ) . This result was consistent with the RNAi screen findings . For an additional test , we examined the effects of combining a heterozygous Csk mutation and a heterozygous cacophony ( cac ) mutation . The cac gene encodes the α1a subunit of the Drosophila CaV2 calcium channel [40] . Cac is a key target of NMJ homeostatic regulation . Partial loss-of-function point mutations in cac , such as cacS , impair synaptic homeostasis [9 , 10 , 22 , 41] . Additionally , prior studies have documented strong heterozygous genetic interactions between cacS and other mutations that disrupt synaptic homeostasis [19 , 30] . Consistent with these prior studies , the NMJs of GluRIIA; cacS/+ heterozygotes showed a partial impairment of their homeostatic response ( Fig 1E ) . Heterozygous Cskj1D8/+ NMJs showed no significant defect in synaptic homeostasis ( Fig 1E ) . However , when cacS/+ and Cskj1D8/+ mutant alleles were combined in a double heterozygous condition , the NMJs were completely unable to enhance quantal content in response to the GluRIIASP16 mutation–in fact QC was slightly depressed compared to the non-GluRIIA mutant control ( Fig 1E ) . These data indicate a strong genetic interaction exists between cac and Csk for proper homeostatic compensation . Potentiation of Cac/CaV2 function and enhancement of presynaptic calcium influx are critical for the rapid induction synaptic homeostasis [10 , 22] . Therefore , we needed to determine whether the Csk gene product also participates in this induction process–or if Csk primarily functions to maintain homeostatic signaling at the NMJ throughout development . Application of the glutamate receptor blocker PhTox to wild-type NMJs for 10 minutes is sufficient to impair glutamate receptor function , causing an acute diminishment in quantal size and a rapid , robust compensatory increase in quantal content ( Fig 1F ) [10] . Interestingly , rapid homeostatic plasticity remained fully intact when we applied PhTox to Csk mutant NMJs ( Fig 1F ) . We conclude that Csk serves to maintain synaptic homeostasis throughout development , but appears to be dispensable for its acute induction . Our data suggested that Csk is required over developmental time for synaptic homeostasis to be executed properly . This could mean that Csk-mediated signaling pathways directly affect homeostatic regulation of neurotransmitter release throughout the course of development . However , it could also mean that depressed Csk levels disrupt homeostatic compensation and neurotransmission indirectly , by negatively impacting other biological processes such as NMJ growth . To distinguish between these possibilities , we turned to immunofluorescence microscopy . First we quantified NMJ growth . We co-stained with antibodies that recognize the presynaptic vesicle protein Synapsin ( anti-Syn ) [42] and the postsynaptic PSD-95 scaffold Discs large ( anti-Dlg ) [43] and used the staining patterns to count NMJ boutons; then we normalized those counts per unit of muscle area . By these measures , NMJ growth was not significantly changed in GluRIIA; Csk mutant NMJs versus wild-type control NMJs ( Fig 2A and 2B and 2E; see S1 Fig for all quantifications of bouton number , muscle size , and bouton count per unit muscle area ) . Therefore , severely impaired synapse growth is unlikely to be responsible for the defects in homeostatic plasticity of GluRIIA; Csk mutants . However , we note that bouton number per unit muscle area was decreased in segment A3 , synapse 6/7 of GluRIIA; Csk double mutants versus Csk single mutants ( Figs 2E and S1C ) . Therefore , defective NMJ development could play some role in impaired homeostatic compensation . To probe this issue further , we examined synaptic active zones . We counted synaptic active zones after immunostaining with antibodies that recognize the presynaptic ELKS/CAST protein Bruchpilot ( anti-Brp ) [44] and the postsynaptic GluRIII glutamate receptor subunit [35] . For larval segment A2 synapse 6/7 , we did not observe any significant changes in total active zone number at GluRIIA; Csk mutant NMJs compared to GluRIIA controls ( Fig 2C–2C” and 2D–2D” and 2F ) . For larval segment A3 synapse 6/7 , GluRIIA; Csk mutant NMJs displayed a slight ( but statistically insignificant ) increase in active zone number compared to GluRIIA controls ( Fig 2F , p = 0 . 07 , Student’s T-Test ) . Therefore , the defect in homeostatic plasticity cannot be explained a decrease in the number of synaptic active zone sites . Next , we examined baseline neurotransmission at Csk mutant NMJs . Compared to wild-type NMJs , neither spontaneous miniature amplitudes ( mEPSPs ) , nor evoked excitatory postsynaptic potentials ( EPSPs ) , nor quantal content ( QC ) were different at homozygous Cskc04256 NMJs or heteroallelic Cskc04256/Cskj1D8 NMJs ( S1 Table , Fig 3A and 3B ) . This result suggested that the electrophysiological defects we measured in GluRIIA; Csk double mutants ( Fig 1C and 1D ) were specifically due to problems in maintaining synaptic homeostasis . We conducted a deeper characterization of baseline Csk electrophysiology . We previously demonstrated that Csk/+ heterozygous mutations interact genetically with CaV2 cacS/+ heterozygous mutations to block synaptic homeostasis ( Fig 1E ) . Therefore , we probed the sensitivity of Csk mutants to differential levels of extracellular Ca2+ . First , we conducted failure analyses in extremely low calcium ( 0 . 14 mM ) [10 , 11] . In low calcium , a significant proportion of evoked presynaptic events fail to release vesicles , resulting in stimulus artifacts , but no discernable EPSPs . Probability of vesicle release is low and quantal content is estimated as QC = ln ( [#trials/#failures] ) [45] . By failure analysis , we found no effect on baseline probability of release for Csk mutant NMJs ( Fig 3C and 3D ) . We also recorded from Csk mutant NMJs over a range of calcium concentrations . We also found no significant change in the calcium cooperativity of release ( Fig 3E ) , suggesting that the calcium sensing machinery in the presynaptic cleft is functional in Csk mutants . Finally , we probed the presynaptic vesicle pool and neurotransmitter release properties by challenging NMJs with a lengthy , high frequency stimulus paradigm ( 10 Hz x 6000 pulses = 10 minutes ) in high extracellular calcium ( 2 mM ) . Under this condition , synapses with an elevated quantal content and release pool–as is the case with GluRIIA NMJs – significantly deplete their evoked responses over the course of such a train ( Fig 3F and 3G ) [10] . By contrast , GluRIIA; Csk double mutant NMJs did not deplete their responses to the same degree as GluRIIA NMJs ( Fig 3F and 3G ) . By the end of the train , there was no excess depletion in GluRIIA; Csk double mutant NMJs compared to wild type or Csk mutant controls ( Fig 3F and 3G ) . Importantly , after a short recovery time following the train ( 30 sec ) , all synapses analyzed displayed fully restored release ( Fig 3G ) . Our data for Csk NMJs and for GluRIIA; Csk NMJs resemble prior examinations of NMJs that have an impaired long-term maintenance of homeostatic plasticity ( e . g . [31] ) . Collectively , the immunostaining and electrophysiology data suggest a specific role in the long-term maintenance of synaptic homeostasis for Csk ( Figs 1–3 ) . The canonical function of Csk family members is to phosphorylate a C-terminal tyrosine residue in Src family kinases ( SFKs ) [46] . This phosphorylation event causes SFKs to assume an auto-inhibitory conformation , and it prevents them from phosphorylating downstream targets [46] . The negative Csk-SFK regulatory relationship is conserved in Drosophila . In vivo and in vitro assays have convincingly demonstrated that Drosophila Csk phosphorylates the Drosophila SFKs Src42A and Src64B at conserved tyrosines [37] . As a result , Drosophila Csk inhibits SFKs in processes such as cell growth and apoptosis [37 , 38] . It is plausible that Csk and SFKs employ the same type of regulatory relationship at NMJs . Indeed , by in situ hybridization ( Berkeley Drosophila Genome Project data ) , enhancer trap analyses , and RNA sequencing analyses , these non-receptor tyrosine kinase genes have been shown to be expressed in many tissues in developing embryos and larvae–including larval central nervous systems and carcasses/musculature [47–53] . Based on those prior findings , the homeostatic defects caused by loss-of-function Csk mutations could be due to overactive SFK function . To test the possibility that overactive SFKs impair synaptic homeostasis , we acquired UAS-SFK transgenes: UAS-Src64BUY1332 [54] , UAS-Src42AWT [37] , and UAS-Src42AYF [37] . The UAS-Src64BUY1332 and UAS-Src42AWT transgenes are wild type . The UAS-Src42AYF transgene is constitutively active , as it lacks the inhibitory tyrosine residue 511 ( Y511F ) that is phosphorylated by Csk [37 , 55] . By combining UAS-SFK expression with a homozygous GluRIIASP16 mutation , and tissue specific GAL4 drivers , we examined how transgenic SFKs could impact NMJ homeostasis . Expression of wild-type UAS-Src64BUY1332 in the muscle completely blocked homeostatic compensation ( Fig 4A ) . By contrast , neuronal expression of wild-type UAS-Src64BUY1332 had no effect on the NMJ’s homeostatic capacity ( Fig 4A ) . The effects of Src42A overexpression were similar to Src64B . Driving either the wild-type UAS-Src42AWT transgene or the constitutively active UAS-Src42AYF transgene in the muscle impaired synaptic homeostasis ( Fig 4B ) . By contrast , neuronal expression of UAS-Src42AYF had no inhibitory effect ( Fig 4B ) . Our results are consistent with the possibility that a muscle-specific Csk/SFK regulatory interaction gates synaptic homeostasis . One qualification is that SFK overexpression had to be limited in order to circumvent embryonic and early larval lethality ( see Materials and Methods and S1 Table for exact conditions for each SFK transgene ) . As a result , this experimental maneuver may have occluded potential Csk-SFK regulatory effects in neurons . For a second approach , we checked if partial SFK loss could suppress the GluRIIA; Csk impairment of synaptic homeostasis . We acquired loss-of-function mutations in Src42A and Src64B [56–60] . To circumvent lethality , we analyzed heterozygous conditions . The heterozygous Src42Ak10108/+ condition increased average evoked amplitudes and QC in a GluRIIA; Csk double mutant background ( S1 Table ) . However , the QC effect was not statistically significant ( Fig 4C ) . By contrast , the heterozygous Src64BKO/+ mutation did partially restore homeostatic capacity to GluRIIA; Csk double mutant NMJs ( Fig 4C and 4E ) . Additionally , control experiments showed that Src64BKO/+ caused no alterations in baseline neurotransmission on its own ( S1 Table ) ; nor did GluRIIA; Src64BKO/+ NMJs significantly differ in homeostatic capacity compared to GluRIIA mutant NMJs ( Fig 4D ) . Collectively , our SFK misexpression and mutant data support the idea that Csk and SFKs play antagonistic roles in the execution of homeostatic plasticity . This is consistent with findings in other organisms and Drosophila tissues . The SFK transgenic expression experiments suggested that a Csk-SFK regulatory process could direct the long-term maintenance of homeostatic compensation from the muscle . However , it is also possible that Csk serves regulatory functions in other tissues through other signaling modalities . Therefore , we performed transgenic Csk expression and rescue experiments . To do this , we generated UAS-Csk-YFP Drosophila stocks ( see Materials and Methods ) , and we expressed Csk-YFP under GAL4 control . In a wild-type genetic background , concurrent muscle and nerve expression of Csk-YFP caused slight electrophysiological abnormalities , including a slight increase in quantal size ( Fig 5A ) , but decreases in evoked postsynaptic excitation ( Fig 5B ) and quantal content ( Fig 5C ) . We imaged concurrent muscle- and nerve-expressed Csk-YFP in unfixed , filleted larvae by fluorescence microscopy . We also used an anti-GFP antibody to stain fixed , filled larvae . In both cases , we observed a strong , punctate cytoplasmic Csk-YFP signal in the muscle ( Fig 5D and 5D’ ) . By immunostaining we could detect low levels of Csk-YFP clustered at the NMJ ( Fig 5D’ and 5D” ) and high amounts of Csk-YFP concentrated at the junction between muscles of adjacent abdominal segments . Next , we expressed Csk-YFP separately in muscles and neurons . By anti-GFP immunostaining , we saw that muscle-specific Csk-YFP protein spread throughout the muscle and clustered at the postsynaptic NMJ ( Fig 5E–5E” ) and muscle attachment sites ( Fig 5E ) . By contrast , we only observed neuronally expressed Csk-YFP protein in the central nervous system ( Fig 5F” ) . We did not observe neuronal-specific Csk-YFP protein at the presynaptic NMJ terminal ( Fig 5F–5F” ) . We further examined the synaptic localization of muscle expressed Csk-YFP by co-staining with monoclonal antibodies against the presynaptic ELKS/CAST active zone marker , Bruchpilot ( Brp , Fig 5G ) , the GluRIIA glutamate receptor subunit ( GluRIIA , Fig 5H ) , and the neural cell adhesion molecule Fasciclin II ( FasII , Fig 5I ) . Not surprisingly , the domain of postsynaptic Csk-YFP staining was largely separate from the region of presynaptic Brp staining ( Fig 5G ) . It also overlapped to only a minor degree with signal from the GluRIIA antibody ( Fig 5H ) . Conversely , there was considerable colocalization between the Csk-YFP and FasII signals , although the two domains did not overlap completely ( Fig 5I ) . These results suggest that Csk-YFP does not localize immediately at sites of synaptic transmission , but has a more periactive zone localization , similar to that previously documented for FasII [61] . We tested whether tissue specific expression of Csk-YFP could restore homeostatic plasticity to GluRIIA; Csk double mutant NMJs . Surprisingly , either pre- or postsynaptic expression of Csk-YFP was sufficient to restore homeostatic compensation ( Fig 5J ) . By contrast , control recordings of UAS-Csk-YFP; GluRIIA; Csk NMJs lacking GAL4 drivers still had completely blocked synaptic homeostasis ( Fig 5J ) . Finally , despite baseline neurotransmission defects associated with Csk-YFP misexpression ( Fig 5A–5C ) , synaptic homeostasis was robust when UAS-Csk-YFP was expressed both pre- and postsynaptically in the context of glutamate receptor loss ( Fig 5K ) . Our rescue experiments employed exogenous Csk-YFP . To probe the tissue specificity of endogenous Csk function , we knocked down Csk gene expression by RNAi . In the RNAi screen , when we knocked down Csk gene expression simultaneously in the pre- and postsynaptic compartments , synaptic homeostasis was completely blocked ( Fig 1A ) . Here we found that postsynaptic knockdown alone partially attenuated synaptic homeostasis ( Fig 5L ) . By contrast , presynaptic Csk knockdown alone left synaptic homeostasis intact ( Fig 5L ) . Taken together the SFK overexpression experiments ( Fig 4 ) , and the tissue-specific Csk knockdown results suggest that postsynaptic ( endogenous ) Csk may be more important than presynaptic Csk in executing synaptic homeostasis . However , it is also true that exogenous Csk can support NMJ homeostatic plasticity when expressed either in the muscle or pan-neuronally . It is puzzling that Csk-YFP is sufficient to gate homeostatic plasticity from either nerve or muscle . One possibility is that Csk controls the release of a homeostatic factor into the synaptic cleft , and this factor is capable of supporting homeostatic function regardless of the tissue of origin . This hypothetical situation would be similar to that described for the secreted homeostatic factor Endostatin , which drives the induction of synaptic homeostasis when expressed either presynaptically or postsynaptically [19] . However , unlike Endostatin , Csk is not required for the rapid induction of synaptic homeostasis ( Fig 1F ) . A second possibility is that Csk limits the release or expression of unknown NMJ factors that could dampen the NMJ homeostat . We attempted to identify NMJ proteins that could be responsive to levels of Csk activity . By immunostaining , we found that Csk mutant NMJ elaboration and bouton expansion were normal ( Fig 2 ) . However , we did note an abnormality: An anti-horseradish peroxidase ( anti-HRP ) antibody which marks neuronal membranes revealed a considerable amount of anti-HRP-positive staining beyond the normal nerve terminal domain at Csk mutant NMJs ( Fig 6 ) . Low-level debris shed from the presynaptic nerve is associated with normal synaptic growth processes [62] . By contrast , excess debris is thought to result from changes in synaptic activity and/or a failure of synaptic and cellular mechanisms to clear debris [62] . Additionally , the disruption of proteins that regulate synapse remodeling can induce small-diameter presynaptic membrane protrusions at the NMJ [63] . For Csk mutant NMJs , it was possible that the abnormal HRP-staining phenotype reflected aberrant synaptic activity , remodeling , or plasticity . Any of these parameters could be relevant to the homeostatic defects we had observed . We conducted additional immunostaining . One molecule we considered for further study was the Neural Cell Adhesion Molecule ( NCAM ) ortholog Fasciclin II ( FasII ) because of the distinctive anti-HRP phenotype and partial co-localization with Csk-YFP ( Fig 5I ) . FasII is one of the glycoproteins recognized by anti-HRP antibodies [64] , and misexpression of FasII can cause abnormalities in synapse development [65] . FasII is present both pre- and postsynaptically; it forms homophilic interactions that span the synaptic cleft , and it is important for multiple forms of synaptic plasticity [66 , 67] . FasII levels at the NMJ are regulated by Mitogen-Activated Protein Kinase/Extracellular Signal-Regulated Kinase ( MAPK/ERK ) [68] . MAPK/ERK is a known target of SFK activity , providing a potential molecular link between Csk and FasII [69–72] . Finally , in vertebrate neurons , many cell adhesion molecules ( CAMs ) gate forms of homeostatic plasticity , possibly as transducers of trans-synaptic signals [73] . Taken together , these facts make FasII a candidate effector of homeostatic Csk activity . We examined FasII protein at Csk mutant NMJs by immunofluorescence . We employed an anti-FasII monoclonal antibody , 1D4 ( University of Iowa , Developmental Studies Hybridoma Bank ) [74] . We also co-stained with anti-HRP and anti-Dlg . We used anti-Dlg staining to define the synapse boundary , and we observed and quantified “synaptic” and “extra-synaptic” anti-FasII staining . We defined staining that extended beyond the Dlg boundary to be extra-synaptic . At wild-type NMJs , extra-synaptic HRP-positive , FasII-positive puncta were present at low levels ( Fig 6A and 6F ) , but in Cskc04256 and Cskc04256/Cskj1D8 NMJs , the relative proportion these extra-synaptic puncta was greatly increased ( Fig 6B and 6C and 6F ) . To check if this staining abnormality was specific to Csk gene function , we attempted to phenocopy it using RNAi . We observed excess extra-synaptic anti-FasII staining when we knocked down Csk function globally ( Tubulin-Gal4 >> UAS-Csk[RNAi] ) ( Fig 6D and 6F ) or in the glia ( Nrv2-Gal4 >> UAS-Csk[RNAi] ) ( Fig 6E and 6F ) . Interestingly , we did not observe the same excess extra-synaptic anti-FasII staining when we knocked down Csk function either pre- or postsynaptically alone ( Fig 7C–7E ) . The fact that the extra-synaptic anti-FasII pattern could be generated independent of neuronal or muscle knock down of Csk indicated that the extra-synaptic FasII staining phenotype itself was unlikely to be causal or predictive for impaired homeostatic signaling . Rather , it seemed to indicate a role for glial-derived Csk in FasII regulation and synapse development . Synaptic contact sites are the loci most likely affected during homeostatic signaling . Therefore , we also quantified synaptic levels of FasII . Cskc04256/Cskj1D8 NMJs showed a statistically significant increase in synaptic FasII compared to wild-type controls ( Fig 7A and 7B and 7F ) . Consistently , knock down of Csk by RNAi in the nerve and muscle or knockdown in muscle only also generated a significant increase in synaptic FasII ( Fig 7C and 7D and 7F ) . By contrast , knock down of Csk in the nerve only caused a slight , but statistically significant decrease in synaptic FasII levels ( Fig 7E and 7I ) . In sum , Csk genetic manipulations that impair or partially impair homeostatic plasticity also result in statistically significant increases in anti-FasII staining at the NMJ . We turned to biochemistry to test if we could identify differences in the expression of individual FasII isoforms in Csk mutants . There exist multiple isoforms of FasII . We generated larval lysates from various genetic backgrounds , ran the lysates on a gradient protein gel , and conducted Western blots using the 1D4 anti-FasII monoclonal antibody . This antibody recognizes known FasII isoforms that contain an intracellular domain , including FasII A-PEST+ ( full ) and FasII A-PEST- ( 29-aa exon excluded ) [75] . For wild-type lysates , we detected two strong FasII A bands , which migrated as predicted for the PEST+ and PEST- 1D4-reactive isoforms ( Fig 7G ) [75] . The gradient gel also revealed a third , previously unreported 1D4-reactive band at a slightly lower molecular weight ( Fig 7G ) . This band could correspond to a poorly characterized , predicted FasII isoform containing an intracellular domain ( FasII H , Accession AHN59326 . 1 ) . It could also be a FasII A degradation product or a non-specific band . We quantified the relative intensity of the 1D4-reactive bands for lysates of four genotypes: 1 ) wild type , 2 ) GluRIIASP16 , 3 ) Cskc04256 , and 4 ) GluRIIASP16; Cskc04256 ( Fig 7G–7I ) . Surprisingly , the intensity of the low molecular weight band was significantly reduced in GluRIIASP16 mutant ( homeostatically competent ) lysates ( Fig 7G and 7I ) . By contrast , in Csk mutant lysates or GluRIIA; Csk mutant ( homeostatically incompetent ) lysates , the low molecular weight FasII band became quite strong , increasing in intensity by > 300% , normalized to lysate actin levels ( Fig 7G and 7I ) . These experiments demonstrate that synaptic FasII levels and isoform-specific levels are responsive to genetic conditions ( GluRIIA and Csk mutations ) that alter the homeostatic capacity of the NMJ . At the NMJ , increased synaptic FasII could contribute to the impairment of homeostatic plasticity . Alternatively , synaptic FasII could have little or no influence on homeostatic plasticity and simply be acting as a secondary biological reporter of Csk-mediated signaling processes . We acquired FasII mutant and overexpression genetic tools to test these ideas . The FasII transposon insertion allele , FasIIEP1462 , is located upstream of the FasII translation start site [76] . This transposon contains a UAS sequence that can be exploited to overexpress FasII from its endogenous locus [76] . We overexpressed FasII at the NMJ by crossing FasIIEP1462 to a Drosophila stock with pre- and postsynaptic GAL4 drivers and a UAS-GluRIII[RNAi] construct to induce a homeostatic challenge [33] ( Fig 8 ) . This cross resulted in an 89 . 5 ± 3 . 4% increase in NMJ FasII levels per unit synapse area relative to wild type ( p < 0 . 001 , Student’s T-Test; Fig 8F–8H ) . It also resulted in a significant impairment of homeostatic compensation ( Fig 8A ) –but not a complete impairment . Next , we employed FasIIEP1462 to overexpress FasII in a tissue-specific manner . Neither presynaptic overexpression of FasII alone , nor postsynaptic overexpression of FasII alone erased synaptic homeostasis in a GluRIIA mutant background ( Fig 8B ) . However , postsynaptic overexpression did cause a partial impairment of synaptic homeostasis that was quantitatively similar to dual-tissue overexpression ( Fig 8B ) . These data are consistent with the possibility that excess postsynaptic FasII is sufficient to impair synaptic homeostasis . We tested if overexpression of individual FasII isoforms could recapitulate the homeostatic defects we observed with overexpression of the entire FasII locus . We acquired three FasII expression transgenes encoding the most well-characterized isoforms: UAS-FasII-A PEST+ , UAS-FasII-A PEST- , and UAS-FasII-C [75] . The two FasII-A isoforms contain extracellular and cytoplasmic domains , and the FasII-C isoform is strictly extracellular and thought to be bound to the membrane by glycosylphosphatidylinositol ( GPI ) linkage [77] . We utilized the transgenes to individually misexpress these three isoforms of FasII in the context of a GluRIII[RNAi] knockdown . Unlike overexpression from the endogenous gene locus using FasIIEP1462 , none of these single misexpression experiments recapitulated the homeostatic impairment ( Fig 8C ) . This result could mean that overexpression of multiple isoforms is needed to block synaptic homeostasis , or it could mean that a different , less well-characterized isoform ( possibly FasII-H ) is responsible for the block in synaptic homeostasis . Csk loss caused dysregulated synaptic FasII ( Figs 6–7 ) , suggesting that FasII may reside downstream of Csk . To characterize the regulatory relationship of Csk and FasII further , we raised Csk levels in the muscle and then tested if this manipulation could ameliorate the defect in synaptic homeostasis caused by FasIIEP14162 overexpression . However , we found that when we overexpressed both UAS-Csk-YFP and FasIIEP14162 in a GluRIIASP16 mutant background , synaptic homeostasis was still impaired ( Fig 8D ) . Our collective data suggest that FasII function resides downstream of Csk function , and excess FasII can overwhelm the ability of Csk to promote synaptic homeostasis ( Figs 6–8 ) . Next we tested if FasII loss-of-function conditions could also cause defects in synaptic homeostasis . Null or strong loss-of-function alleles like FasIIG0293 [78] show lethality at various stages of development , whereas FasIIe76 is a semi-viable , partial loss-of-function mutation [67] . In a GluRIII[RNAi] knockdown background , we examined the electrophysiological effects of the FasIIe76 mutant allele as a hemizygote ( FasIIe76/Y males ) or in combination with the FasIIG0293 allele ( FasIIe76/FasIIG0293 females ) . We also tried knocking down FasII function using RNAi lines , FasIIGD14486 ( Vienna Drosophila Resource Center line , v36351 ) and FasIIJF02918 ( Bloomington Drosophila Stock Center TRiP collection ) . None of these FasII loss-of-function conditions impaired homeostatic compensation ( Fig 8E ) . The most plausible interpretation is that synaptic homeostasis is sensitive to elevated levels of FasII , but not to diminished levels of FasII . Alternatively , the low levels of FasII present at hypomorphic NMJs may be sufficient for homeostatic compensation to occur . Misexpression of FasII could disrupt NMJ development and cause homeostatic defects as a consequence of aberrant NMJ growth . Overexpression of FasII in neurons has been reported to result in axon pathfinding defects , ectopic synapses at some sites , and reduced or abnormal synapses at other sites [76] . Furthermore , concurrent pre- and postsynaptic overexpression of the GPI-linked form of FasII has been reported to induce the production of satellite boutons [79] , and misexpression of FasII in muscle fibers can alter NMJ innervation patterns [65] . Therefore , we needed to determine if the FasII overexpression conditions we used to impair synaptic homeostasis also triggered gross abnormalities in NMJ development . Compared to wild-type synapses , we found no significant changes in bouton number or muscle area when FasIIEP1462 was overexpressed from its endogenous locus during a GluRIII[RNAi]-induced homeostatic challenge ( S1D–S1F Fig ) . This result supports the idea that the homeostatic defects associated with FasII overexpression are not caused by gross NMJ undergrowth . We have shown that the Csk mutant NMJs have altered FasII NMJ expression ( Figs 6 and 7 ) . Additionally , misexpression of FasII partially impairs synaptic homeostasis ( Fig 8 ) ; loss of FasII leaves synaptic homeostasis intact ( Fig 8 ) . Therefore , we tested whether reducing FasII levels could restore homeostatic compensation to GluRIIA; Csk mutant NMJs . GluRIIA; Csk double mutants had completely blocked synaptic homeostasis ( Figs 1 and 9A ) , but remarkably , synaptic homeostasis was restored in the FasIIe76/Y; GluRIIASP16; Cskc04256/J1D8 triple loss-of-function mutant ( Fig 9A–9C ) . This result suggested that homeostatic defects observed at GluRIIA; Csk mutant NMJs could be due to synaptic FasII misexpression . To verify that the FasIIe76 genetic manipulation employed in our electrophysiological analyses depressed FasII protein expression , we conducted immunostaining and Western blotting . FasIIe76 markedly reduced synaptic FasII protein ( Fig 9D–9F ) , and it also muted the expression of all 1D4-reactive FasII bands on Western blots , including the low molecular weight band that is enhanced in Csk loss-of-function mutants ( Fig 9G–9I ) . Finally , we assessed NMJ growth in FasII; Csk double mutants by counting synaptic boutons and measuring muscle size . Absolute bouton number was unchanged relative to wild type ( S1A Fig ) . However , due to a small decrease in muscle size , bouton number per unit muscle area was increased for segment A2 , synapse 6/7 ( S1B and S1C Fig ) . This change persisted whether there was a GluRIIASP16 homeostatic challenge or not ( S1C Fig ) . Together , our data show that FasII loss-of-function mutations can suppress the defects in synaptic homeostatic plasticity caused by Csk mutations . There may be a slight developmental growth component that correlates with this electrophysiological suppression .
While Csk-YFP expression was sufficient to restore homeostatic compensation either pre- or postsynaptically , we favor a model in which postsynaptic Csk plays the predominant role in synaptic homeostasis . This is supported by several observations . First , Csk-YFP was only detected at the NMJ when postsynaptically expressed ( Fig 5 ) . Additionally , when the endogenous function of Csk was examined by RNAi-meditated knockdown , only postsynaptic Csk knockdown led to impaired synaptic homeostasis ( Fig 5 ) and increased FasII levels ( Fig 7 ) . Additionally , postsynaptic FasII overexpression was sufficient to impair homeostatic plasticity ( Fig 8 ) . These combined data suggest that , in the case of FasII-mediated homeostatic processes , postsynaptic Csk is the primary player in synaptic homeostasis . The ability of presynaptic Csk-YFP expression to restore homeostatic competency indicates that presynaptic Csk may also play a role in these processes , likely through a FasII-independent pathway . The identification of homeostatic functions for Csk and SFKs in the muscle expands our understanding of the postsynaptic proteins that regulate Drosophila NMJ neurotransmission through retrograde signaling . A small number of postsynaptic factors have been previously implicated in this process , including the protein scaffold Dystrophin ( Dys ) [80 , 81] , the RhoGAP Crossveinless-c ( Cv-c ) [80] , the nuclear import factor Importin-13 ( Imp13 ) [82] , Calcium/Calmodulin protein Kinase II ( CaMKII ) [83] , Drosophila target of rapamycin ( TOR ) [32] , and S6 Kinase ( S6K ) [32] . In the cases of Dys , Cv-c , and Imp13 , loss of any of these factors in the muscle causes a de novo increase of presynaptic neurotransmitter release [80–82] . This likely means that they are negative regulators of retrograde signaling . However , these factors remain to be fully characterized in the context of a challenge to synapse function , such as glutamate receptor loss . In the cases of TOR and its phosphorylation target , S6K , loss-of-function mutations block the long-term maintenance of homeostatic compensation [32] , and postsynaptic overexpression or activation of these factors instructively induces de novo increases in presynaptic quantal content [32] . As a homeostatic signaling molecule , Csk appears to be distinct from these previously described factors . One difference between TOR/S6K and Csk is that elevated Csk levels do not induce a de novo increase in quantal content ( Fig 5C ) . Moreover , transgenic Csk-YFP only induces increases in vesicle release in the context of a homeostatic challenge , like glutamate receptor loss ( Fig 5K ) . Another difference is that Csk is capable of conferring homeostatic competence to the NMJ from either the nerve or the muscle ( Fig 5J ) , while factors like TOR and S6K perform their homeostatic functions from the muscle [32] . This result suggests that Csk could be more generally involved in regulating the expression or localization of factors that influence homeostatic plasticity over extended periods of developmental time . One unresolved issue is how exogenous Csk-YFP was able to restore homeostatic capacity to Csk mutant NMJs . It is possible that neuronal Csk and SFKs regulate unknown cellular targets that drive the presynaptic expression of homeostatic plasticity . Some avenues for future experiments are suggested by prior studies . For example , we know from vertebrate neuronal preparations that SFKs are capable of regulating neurite outgrowth , cell shape and cell growth by phosphorylating the guanine exchange factor , Ephexin ( Exn ) [84–87] . At the NMJ , we have previously demonstrated that presynaptic Exn supports the long-term maintenance of homeostatic plasticity in conjunction with CaV2-type calcium channels [30] . Therefore , it is possible that presynaptic Csk can regulate synaptic homeostasis via Exn regulation . Another possible presynaptic Csk/SFK target is the vesicle-associated protein Synapsin ( Syn ) . In rodent neurons , Src phosphorylation of Syn diminishes the readily releasable pool ( RRP ) of synaptic vesicles , and expression of a non-phosphorylatable version of Synapsin ( Y301F ) increases the size of the RRP [88 , 89] . Since RRP size increase is a key determinant in homeostatic compensation at the Drosophila NMJ [20 , 23] , Csk mutations might lead to excessive SFK-mediated phosphorylation of Drosophila Syn and a failure to increase RRP size and QC when the synapse is homeostatically challenged . Such a model could explain our finding that GluRIIA; Csk double mutant NMJs do not deplete evoked responses to the same degree as GluRIIA mutant NMJs when challenged with a high frequency stimulus train ( Fig 3F–3G ) . We know very little about how synapses maintain normal levels of function for extended periods of developmental time . When considering the Drosophila NMJ , it is helpful to compare and contrast the long-term maintenance of synaptic homeostasis with its acute induction . Induction can occur on a timescale of minutes; it does not require protein translation , and it occurs even when the presynaptic nerve is severed from the neuronal soma [10] . By contrast , the long-term maintenance of synaptic homeostasis requires protein translation-activating factors in the muscle ( TOR/S6K ) [32] , cytoplasmic signaling molecules in the neuron ( the Rho-GEF Ephexin ) [30] , and a neuronal transcription factor ( Gooseberry ) [31] . These factors seem to act in concert to consolidate increases in QC in response to a homeostatic challenge . To this short list , we add C-terminal Src kinase , which can support homeostatic compensation from either the neuron or the muscle . We also add the misexpression of Src42A , Src64B , and FasII as factors that impair homeostatic outputs . One caveat to our conclusion that Csk acts as a factor that controls only the long-term consolidation of synaptic homeostasis is that the Csk mutant conditions used for this study were not null . Csk null mutants arrest well before the third instar larval stage . Therefore , our data cannot formally rule out a role for Csk in the acute induction of homeostatic plasticity . How synapses work to stabilize outputs for extended periods of developmental time could be relevant for understanding neurological disorders that manifest after decades of life . An interesting example that has direct parallels to our experimental system is the neuromuscular disorder myasthenia gravis ( MG ) . Autoimmune inhibition of human NMJ acetylcholine receptors ( AChRs ) is one cause of MG . In a phenomenon that is similar to the homeostatically challenged Drosophila NMJ , electrophysiological recordings from myasthenic muscle show that these NMJs have a reduced quantal size and a compensatory increase in QC [90 , 91] . It is possible that the cellular events that drive the increase in QC at myasthenic NMJs serve to quell disease manifestations until later in life . Interestingly , the NCAM CD56 has been shown to be decreased in cells treated with corticosteroids , which is a common method to combat MG [92] . FasII forms homophilic trans-synaptic complexes [66 , 67 , 77] . Therefore , FasII is an intriguing candidate to modulate retrograde signaling at the NMJ . To date , examples of CAMs ( or CAM complexes ) that have been implicated in synaptic homeostasis include N-cadherin/β-catenin [93 , 94] , β3 integrins [95 , 96] , Ephrin ligands/Eph receptors [30 , 97 , 98] , class I major histocompatibility complex proteins ( MHC-1 ) [99] , and α-neurexins [100] . Some of these CAMs are hypothesized to serve as scaffolds that draw other molecules to the synapse; in turn , the recruited molecules could modulate neurotransmission . [73] . How might synaptic FasII/NCAM impinge upon homeostatic signaling in the context of Csk loss ? Prior work has demonstrated that pre- and postsynaptic FasII can exert effects on synaptic stabilization [67] . Temporary increases in muscle FasII in particular can result in novel synaptic contacts that are stable [65] . In principle , it is possible that our Csk and FasII genetic manipulations spurred a small number of inappropriate synaptic connections that occluded the homeostatic capacity of the NMJ . A second potential explanation for our data is that FasII impairs the function of a known homeostatic regulator . At the NMJ , the majority of identified homeostatic factors function in the presynaptic neuron [101] . These include regulators of presynaptic vesicle release , ion channels , cytoplasmic signaling molecules , and key components of the active zone [101] . Among these factors , one potential target of FasII/NCAM regulation is the Drosophila Eph receptor ( Eph ) [30] , upstream of Exn . Compelling parallels exist between Drosophila and mammalian data regarding the function of Eph receptors in synaptic plasticity . Mammalian NCAM physically interacts with EphA Receptors ( EphAs ) [102 , 103] , and an NCAM/EphA interaction was recently reported to dictate excitatory/inhibitory balance in the mouse prefrontal cortex [103] . At the Drosophila NMJ , it was previously reported that the Eph receptor functions in conjunction with the cytoplasmic guanine exchange factor Exn and CaV2-type calcium channels to homeostatically modulate presynaptic neurotransmitter release [30] . Since Eph and Exn also dictate the long-term maintenance of NMJ homeostasis [30] , this opens the possibility that trans-synaptic FasII acts as a modulator of Eph . Other retrograde factors proposed to function in the synaptic cleft for homeostatic compensation are the BMP ligand Glass Bottom Boat ( Gbb ) [15 , 104] and the Drosophila homolog of Endostatin [19] . Endostatin is produced when the Collagen XV/XVIII homology factor Multiplexin is cleaved by matrix-metalloproteinases ( MMPs ) or cysteine cathepsins [105] . Interestingly , in mammalian systems , increased NCAM expression leads to a decrease in the secretion of several MMPs [106] . This relationship has not been examined directly in Drosophila , but FasII and MMPs are both required for proper motor neuron fasciculation , axon guidance , and target recognition [107–112] , suggesting that a regulatory relationship between these molecules may be present in Drosophila as well . Future experiments will be needed to determine precisely how increased FasII expression at the NMJ results in reduced homeostatic capacity .
Drosophila stocks carrying various mutant alleles , GAL4 drivers , or UAS-driven transgenes were used for this study . Stocks were either obtained from the Bloomington Drosophila Stock Center ( BDSC , Bloomington , IN ) , the Vienna Drosophila RNAi Center ( VDRC , Vienna , Austria ) , or from researchers who generated them . Mutant alleles include: Cskc04256 [36] , Cskj1D8 [39] , Src42Ak10108 [59] , FasIIEP1462 [76] , FasIIG0293 [78] , FasIIe76 [77] , Src64BKO [113] . GAL4 drivers include: elaV ( C155 ) -GAL4 [114] , OK371-GAL4 [107 , 115] , Scabrous-GAL4 [43] , BG57-GAL4 [43] , MHC-GAL4 [67] , MyoD-GAL4 [116] . UAS transgenes include: UAS-Src64BUY1332 [54] , UAS-Src64B . C ( J . M . Dura and J . Cooper donation to BDSC ) , UAS-Src42AYF [37] , UAS-Src42AWT [37] . UAS-RNAi lines from VDRC [117] include: UAS-Csk[RNAi] ( VDRC# 32877 , CskGD9345 ) and UAS-FasII[RNAi] ( VDRC# 36351 , FasIIGD14486 ) . UAS-RNAi lines from BDSC include: UAS-FasII[RNAi] ( FasIIJF02918 ) . We constructed UAS-GluRIII[RNAi] [33] . To construct UAS-Csk-YFP transgenic lines , we PCR amplified the PH isoform of Csk from cDNA clone LP09923 from the Drosophila Genomics Resource Center ( DGRC , Bloomington , IN ) , and subsequently cloned the amplicon into a pUAST vector with a C-terminal YFP tag , using the Drosophila Gateway Vector Collection ( T . Murphy , Carnegie Institute of Washington ) . Transgenic animals were made by injecting DNA into embryos using standard procedures ( BestGene , Inc . , Chino Hills , CA ) . Fruit flies were reared in chambers with temperature control; experimental and control animals were raised in parallel under identical conditions . We conducted UAS-Src64B and UAS-Src42A ( UAS-SFK ) overexpression experiments by using elaV ( C155 ) -GAL4 or OK371-GAL4 for presynaptic expression and BG57-GAL4 or MHC-GAL4 for postsynaptic expression . In all cases for UAS-SFK expression , the conditions were lethal when animals were raised at 29°C , but lowering the rearing temperature was sufficient to circumvent lethality and allow for electrophysiological analyses . The specific UAS-SFK/GAL4 combinations and rearing temperatures that we used for analysis are found in S1 Table . Wandering third instar larvae were selected for electrophysiological recordings . Sharp electrode recordings were taken from muscle 6 of abdominal segments 2 and 3 , as previously described [10 , 30 , 118] . Larvae were dissected in a modified HL3 saline with the following components ( and concentrations ) : NaCl ( 70 mM ) , KCl ( 5 mM ) , MgCl2 ( 10 mM ) , NaHCO3 ( 10 mM ) , sucrose ( 115 mM = 3 . 9% ) , trehalose ( 4 . 2 mM = 0 . 16% ) , HEPES ( 5 . 0 mM = 0 . 12% ) , and CaCl2 ( 0 . 5 mM , unless otherwise indicated ) . Data were collected using Axopatch 200B or Axoclamp 900A amplifiers ( Molecular Devices , Sunnyvale , CA ) , digitized using a Digidata 1440A data acquisition system ( Molecular Devices ) , and recorded with pCLAMP 10 acquisition software ( Molecular Devices ) . For presynaptic nerve stimulation , a Master-8 pulse stimulator ( A . M . P . Instruments , Jerusalem , Israel ) and an ISO-Flex isolation unit ( A . M . P . Instruments ) were utilized to deliver 1 ms suprathreshold stimuli to the appropriate segmental nerve . The average spontaneous miniature excitatory postsynaptic potential ( mEPSP ) amplitude was quantified by measuring the amplitude of approximately 100–200 individual spontaneous release events per NMJ ( Mini Analysis , Synaptosoft , Fort Lee , NJ ) . The average per-NMJ mEPSP amplitudes were then averaged for each genotype . The average evoked EPSP amplitude was calculated for each NMJ . Quantal content was determined for each recording by calculating the ratio of average EPSP and average mEPSP amplitudes . Quantal contents were calculated for each recording and then averaged across NMJs of the indicated genotypes . Where indicated in S1 Table , quantal contents were corrected for non-linear summation as described [119] . Long trains of stimuli were analyzed using an auto-analyze function in Mini Analysis . Third instar larvae were filleted in HL3 saline . Dissected animals were fixed for 3 minutes in Bouin’s fixative ( Ricca Chemical Company , Arlington , TX ) , washed using standard procedures , and incubated in primary antibodies at room temperature for two hours . This was followed by additional washes and another two-hour incubation in secondary antibody at room temperature . Staining was performed using the following primary antibodies: mouse anti-Synapsin ( 3C11 ) 1:50 [42] ( Developmental Studies Hybridoma Bank , University of Iowa–DSHB – deposited by Buchner , E . ) ; rabbit anti-Dlg 1:30 , 000 [43]; mouse anti-Brp ( nc82 ) 1:250 [44] ( deposited to DSHB by Buchner , E . ) ; mouse anti-FasII ( 1D4 ) 1:900 [74] ( deposited to DSHB by Goodman , C . ) ; mouse anti-GluRIIA ( 8B4D2 ) 1:500 [120] ( deposited to DSHB by Goodman , C . ) . The following fluorophore-conjugated antibodies were also used ( Jackson ImmunoResearch Laboratories , Inc . , West Grove , PA ) : goat anti-mouse-488 1:1000 ( DyLight ) ; goat anti-rabbit-549 1:2000 ( Dylight ) ; goat anti-HRP-TRITC 1:1000 , Alexa-647 goat anti-HRP 1:500 . Larval preparations were mounted in Vectashield ( Vector Laboratories , Burlingame , CA ) and imaged at room temperature using Zen software on a Zeiss 700 LSM mounted on an Axio Observer . Z1 using an EC Plan-Neofluar 40X Oil DIC Objective ( aperture 1 . 30 ) or an EC Plan-Apochromat 63x Oil DIC Objective ( aperture 1 . 40 ) ( Carl Zeiss Microscopy , Jena , Germany ) . For each experiment , experimental and control larval preps were stained in the same container , mounted on the same slide , imaged using identical acquisition settings , and analyzed using the same procedure and thresholds . Bouton and active zone numbers were quantified semi-automatically using the ‘Spots’ function in Imaris x64 v7 . 6 . 0 ( Bitplane , Zurich Switzerland ) . Boutons were counted using the anti-Synapsin channel with the XY diameter set at 3 μm . Active Zones were counted using the anti-Brp channel with an XY diameter of 0 . 3 μm . The threshold was adjusted so that each bouton/active zone was recognized once . Any errors in automated counting were corrected by hand to arrive at the final value . FasII levels and localization were assessed using ImageJ 1 . 48s/Java 1 . 6 . 0_24 ( 64-bit ) with Fiji plugins . FasII levels were assessed as follows: Z-stack images were compressed using the maximum projection function; regions of interest ( ROIs ) were hand drawn to exclude any non-synaptic structures in the image; a minimum threshold was set for each channel to eliminate background fluorescence and held consistent within each experiment; the Measure function was used to assess fluorescence intensity and area for each channel ( FasII ( 488 ) , Dlg ( 549 ) , HRP ( 647 ) ) ; total FasII and Dlg intensities were normalized to the synaptic area of the HRP channel to control for variation in synaptic size . Extra-synaptic FasII levels were determined as follows: Z-stack images were compressed using the maximum projection function; a mask was generated based on the Dlg channel; this mask was used to generate ROIs that defined the synaptic area and the extra-synaptic area; a third ROI was hand drawn to exclude any non-synaptic structures in the image; this third ROI was combined with the DLG mask-based ROIs to generate the final synaptic and extra-synaptic ROIs used for analysis; FasII fluorescence intensity was measured in these ROIs; FasII and Dlg fluorescence intensity and HRP fluorescence area were also measured at each synapse independent of the ROIs; FasII fluorescence intensity/HRP synaptic area was calculated for the whole synapse , the synaptic region , and the extra-synaptic region; the percent extra-synaptic FasII was calculated based on these values . SDS-PAGE was performed using the Novex NuPAGE SDS-PAGE system with 4%-12% Bis-Tris gels run at 125 V for 5 hours to achieve separation of individual FasII bands . Transfer to nitrocellulose membrane ( Whatman , Dassel , Germany ) was performed using a Trans-Blot-SD Semi-Dry Transfer Cell ( Bio-Rad , Hercules , CA ) . Blocking was performed in 5% BSA for FasII blots or 5% milk for actin blots in 1X PBS with 0 . 1% Tween 20 . Primary antibodies were obtained from the DSHB: mouse anti-FasII ( 1D4 ) 1:100 and mouse anti-actin ( JLA20 ) 1:1000 . Horseradish peroxidase-conjugated goat anti-mouse secondary antibody ( Jackson ImmunoResearch Laboratories , Inc . , West Grove , PA ) was used at 1:1000 for FasII blots and 1:5000 for actin blots . All antibodies were diluted in blocking buffer . Blots were developed with SuperSignal West Pico Chemiluminescent Substrate ( Thermo Scientific , Waltham , MA ) and imaged with Amersham Hyperfilm ECL film ( GE Healthcare Limited , Buckinghamshire , UK ) . Band intensity was quantified using ImageJ . Statistical significance was assessed either by Student’s T-Test with direct comparison between an experimental data set and a control data set , or one-way ANOVA with Tukey’s post-hoc test across multiple data sets , as appropriate . Specific p value ranges are given in the figure legends , with * p < 0 . 05 , ** p < 0 . 01 , and *** p < 0 . 001 marked as significant . Some p values that may trend toward significance ( p < 0 . 1 ) are also indicated . The values reported in text or plotted on bar graphs are mean ± SEM , with n values for experimental trials placed on a bar . Raw values for electrophysiological data–including n values for control and exact experimental genotypes–are given in S1 Table .
|
Homeostasis is a fundamental topic in biology . Individual cells and systems of cells constantly monitor their environments and adjust their outputs in order to maintain physiological properties within ranges that can support life . The nervous system is no exception . Synapses and circuits are endowed with a capacity to respond to environmental challenges in a homeostatic fashion . As a result , synaptic output stays within an appropriate physiological range . We know that homeostasis is a fundamental form of regulation in animal nervous systems , but we have very little information about how it works . In this study , we examine the fruit fly Drosophila melanogaster and its ability to maintain normal levels of synaptic output over long periods of developmental time . We identify new roles in this process for classical signaling molecules called C-terminal Src kinase , Src family kinases , as well as a neuronal cell adhesion molecule called Fasciclin II , which was previously shown to stabilize synaptic contacts between neurons and muscles . Our work contributes to a broader understanding of how neurons work to maintain stable outputs . Ultimately , this type of knowledge could have important implications for neurological disorders in which stability is lost , such as forms of epilepsy or ataxia .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"nervous",
"system",
"electrophysiology",
"neuroscience",
"homeostatic",
"mechanisms",
"animals",
"animal",
"models",
"physiological",
"processes",
"drosophila",
"melanogaster",
"model",
"organisms",
"synaptic",
"plasticity",
"homeostasis",
"bioassays",
"and",
"physiological",
"analysis",
"muscle",
"electrophysiology",
"drosophila",
"nervous",
"system",
"physiology",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"developmental",
"neuroscience",
"electrophysiological",
"techniques",
"insects",
"arthropoda",
"cellular",
"neuroscience",
"immunostaining",
"anatomy",
"synapses",
"neural",
"homeostasis",
"physiology",
"biology",
"and",
"life",
"sciences",
"neurophysiology",
"organisms"
] |
2016
|
C-terminal Src Kinase Gates Homeostatic Synaptic Plasticity and Regulates Fasciclin II Expression at the Drosophila Neuromuscular Junction
|
Macrophages are the first line of defense against pathogens . Upon infection macrophages usually produce high levels of proinflammatory mediators . However , macrophages can undergo an alternate polarization leading to a permissive state . In assessing global macrophage responses to the bacterial agent of Whipple's disease , Tropheryma whipplei , we found that T . whipplei induced M2 macrophage polarization which was compatible with bacterial replication . Surprisingly , this M2 polarization of infected macrophages was associated with apoptosis induction and a functional type I interferon ( IFN ) response , through IRF3 activation and STAT1 phosphorylation . Using macrophages from mice deficient for the type I IFN receptor , we found that this type I IFN response was required for T . whipplei-induced macrophage apoptosis in a JNK-dependent manner and was associated with the intracellular replication of T . whipplei independently of JNK . This study underscores the role of macrophage polarization in host responses and highlights the detrimental role of type I IFN during T . whipplei infection .
Over the past decades , activated macrophages were mainly considered as cells that secrete inflammatory mediators and kill intracellular pathogens . However , studies have now revealed activated macrophages as a continuum of cells with phenotypic and functional heterogeneity [1] , [2] . Schematically , macrophages exposed to the classic activation signals ( lipopolysaccharide ( LPS ) and/or IFN-γ ) polarize into the M1 phenotype and express high levels of TNF , IL-1 , IL-6 , IL-12 , type I IFN , inflammatory chemokines , such as CXCL10 , and inducible nitric oxide synthase . In contrast , M2 macrophages , induced by IL-4 , IL-10 or immune complexes , are characterized by the expression of non-opsonic receptors , arginase , and the absence of proinflammatory cytokines [3] , [4] . Recently , we defined a “common host response” of macrophages to bacterial infections , characterized with an M1 signature and associated with the control of acute infections . However , successful infection by pathogenic intracellular bacteria usually relies on the perturbation or avoidance of the classical M1 proinflammatory activation profile [3] . Recognition of microorganisms by macrophages is mediated by pattern recognition receptors ( PRR ) that bind conserved microbe-associated molecular patterns ( MAMPs ) [5] . PRR engagement by MAMPs activates a major signaling cascade that leads ultimately to the activation of mitogen-activated protein ( MAP ) kinases and the transcription factors NF-κB and IRF3 [5] , [6] . These transcription factors then migrate to the nucleus where they drive the transcription of proinflammatory genes and type I IFN genes , respectively [7] . Type I IFN are responsible for inducing transcription of a subset of genes referred as interferon stimulated genes . Classically , type I IFN transcription is first activated by signals that induce cooperative binding of the transcription factors c-Jun/ATF2 , NF-κB and interferon regulatory factor-3 ( IRF3 ) to the promoter [8] . Following stimulation with viral or bacterial components , the constitutively expressed IRF3 is phosphorylated in the cytoplasm , dimerizes and then translocates in the nucleus to induce the transcription of type I IFN [9] . Once secreted , type I IFN initiates a positive feed-back loop through binding to its receptor IFNAR [8] . IFNAR activates the protein tyrosine kinases JAK1 and JAK2 which phosphorylate STAT1 and STAT2 to further drive the transcription of a large group of IFN inducible genes [10] . Stimulation with Gram-negative bacteria or LPS induces type I IFN , at least partially through Toll-like receptor ( TLR ) 4 [11] . In addition , the intracellular pathogens Shigella flexnerii , Legionella pneumophila and Francisella tularensis induce a potent type I IFN response while non invasive mutants do not [12]–[14] . MAMPs from Gram-positive bacteria are also able to induce type I IFN . Indeed , Listeria monocytogenes triggers type I IFN , probably through bacterial DNA recognition by a cytosolic receptor [15] , [16] . Infection of various cell types with Mycobacterium tuberculosis has also been shown to induce type I IFN [17] . Recently , the extracellular pathogen group B Streptococcus has been shown to induce type I IFN in a TLR-independent manner through intracellular recognition of its DNA [18] . Remarkably , stimulation of macrophages with most of these bacteria and/or bacterial ligands induces M1 polarization , strongly supporting the fact that type I IFN response is a feature of classical activation of macrophages . This point is strengthened by the fact that type I IFN significantly contribute to the cross-talk between the MyD88-dependent and MyD88-independent pathways , enabling full responsiveness to LPS [19] . Here , we have studied and characterized mouse macrophage responses to infection with the facultative intracellular Gram positive bacterium Tropheryma whipplei , the etiologic agent of Whipple's disease [20] . Whipple's disease is a rare systemic disease that associates arthropathy , weight loss and gastrointestinal symptoms [21] but its pathophysiology remains largely unknown . Recently , we provided major insights into the understanding of host immune factors in Whipple's disease , delineating macrophage polarization and apoptosis as critical in the pathophysiology of the disease [3] , [22]–[24] . In this report , we provide evidence that , besides M2 macrophage polarization and apoptosis , T . whipplei induced a robust type I IFN response . This response required bacterial viability and was associated with bacterial intracellular replication . We also observed that T . whipplei induced macrophage apoptosis in a type I IFN- and JNK- dependent manner . These findings reveal an unexpected type I IFN response associated with M2 polarization .
To evaluate gene expression profiles , bone marrow-derived macrophages ( BMDM ) were infected with T . whipplei for 6 h and transcriptional response was examined by microarray analysis . Of the 43 , 379 spotted features , 356 were significantly modulated in response to T . whipplei infection ( P<0 . 01 , Fig . 1A ) . To increase the reliability of our datasets , we considered transcripts as significantly regulated if they showed at least a 2-fold modulation in gene expression levels . We overall identified 59 and 11 genes that were up- and downregulated , respectively . Upregulated genes were assigned to biological process gene ontology ( GO ) categories . Around 50% of them belonged to the immune response GO group ( Fig . 1B ) . These immune response genes could be sub-classified in 4 functional categories . In the first category were genes linked to macrophage polarization and more specifically to M2 polarization ( Fig . 1C ) . Indeed , genes for the prototypal M2 markers interleukin 1 receptor antagonist ( il1rn ) and arginase 2 ( arg2 ) , as well as the M2 chemokines Ccl17 and Ccl22 were induced , while none of the M1 markers were modulated . The second set of the immune response-related genes represented genes related to PRR ( Fig . 1C ) . In this group , 2 genes encoding lectins were markedly induced: clec4e which encodes a C-type lectin , and olr1 , which encodes the lectin-like oxidized low density lipoprotein receptor 1 . In addition , T . whipplei also upregulated the expression of tlr2 , involved in the recognition of Gram-positive bacteria , and that of formyl peptide receptor 2 , encoded by fpr2 , which mediates the chemotactic activity of a variety of pathogen and host-derived peptides . The third group of immune response-associated genes included apoptosis-related genes . Indeed , we found that fas , and tnfsf10 were efficiently induced in BMDM following stimulation with T . whipplei ( Fig . 1C ) . Finally , we isolated a fourth set of immune response-related genes that contained genes involved in the type I IFN response ( Fig . 1C ) . In this group were found the genes encoding Mx1 and Mx2 , which mediate resistance against negative-strand RNA viruses , but also the IFN-stimulated genes ifit1 , ifit2 and ifit3 , also known as isg56 , isg54 and isg49 , respectively . Three other IFN-inducible genes ( irg1 , ifi44 and gbp2 ) were among the most induced genes by T . whipplei . Selected genes were studied by quantitative real time RT-PCR . Upregulation of these exemplary genes was confirmed and statistical analysis revealed a significant correlation between microarray and RT-PCR data ( Table 1 ) . The transcriptional program of BMDM elicited by T . whipplei revealed a marked polarization towards a M2 phenotype . This macrophage functional activation state is characterized by the absence of proinflammatory mediators [3] . We investigated the lack of proinflammatory response by examining the activation of the transcription factor NF-κB and the phosphorylation of MAPK in response to T . whipplei . NF-κB activation was assessed by determining changes in cytoplasmic IκBα protein levels . LPS ( 100 ng/ml ) clearly induced NF-κB activation . Indeed , a transient degradation of IκBα , maximal at 15 min was observed ( Fig . 2A ) . This profile was in agreement with kinetics of RelA translocation in the nucleus ( Fig . 2B ) . In contrast , stimulation with T . whipplei induced a faint IκBα degradation between 1 h and 2 h and IκBα levels increased back normal by 3 h ( Fig . 2A ) . However , RelA translocation was not observed , even by increasing 4 fold the dose of bacteria ( Fig . 2B ) . Nevertheless , the fact that IκBα increased back to initial levels at 3 h suggest that T . whipplei is a weak activator of NF-κB . Besides NF-κB , we assessed MAPK activation in response to T . whipplei . BMDM were stimulated with 100 ng/ml LPS or T . whipplei for 15 min to 3 h and , subsequently , analyzed for phosphorylation of the MAPKs , p38 , Erk1/2 , and JNK . In the first 15–30 min after LPS stimulation , transient phosphorylation of all kinases could be detected ( Fig . 2C ) . In contrast , when BMDM were stimulated with T . whipplei , no phosphorylation of the MAPKs p38 , ERK and JNK could be observed during the 3 hour-time-frame ( Fig . 2C ) . Increasing the doses of T . whipplei had no effect on MAPK activation ( data not shown ) . As BMDM were poorly proinflammatory , it is likely that they allowed T . whipplei replication . Therefore , BMDM were infected with T . whipplei and bacterial uptake and replication was assessed by qPCR . BMDM efficiently internalized T . whipplei as around 6 , 000 bacterial DNA copies were detected after 4 h of infection ( Fig . 3A ) . In the first 3 days , bacterial DNA copy number decreased and started to increase after 6 days and reached around 30 , 000 copies after 12 days ( Fig . 3A ) . These results were further investigated by examining the vacuole containing T . whipplei at day 12 post infection . The great majority of bacteria colocalized with the late phagosome marker lamp1 ( 92%±11% ) ; however , these T . whipplei-containing vacuoles excluded the lysosomal hydrolase cathepsin D ( 20%±15% , Fig . 3B ) . Overall , these results showed that T . whipplei infects BMDM , induces M2 polarization and replicates , at least by interfering with phagosome conversion . Besides M2 polarization , BMDM response profiling to T . whipplei infection revealed a striking induction of type I IFN-inducible genes . Some genes , among which ifnb1 and cxcl10 , which encode respectively IFN-β and the chemokine Cxcl10 , were excluded from the analysis when we applied our criterion; however , ifnb1 was up-regulated 1 . 6 times and cxcl10 4 . 2 times . To further confirm type I IFN induction following T . whipplei infection , we performed time course experiments . Expression of IFN-β mRNA increased to reach a maximal level at 6 h after infection and then was shut off , as revealed by its low expression value at 24 h ( Fig . S1A ) . Consistent with transcriptional data , IFN-β protein was secreted by infected BMDM at 3 h and reached maximal levels 6 h post infection ( Fig . S1B ) , exemplifying the importance of the type I IFN pathway during T . whipplei infection . Induction of IFN-β is thought to depend on the constitutively expressed transcription factor IRF3 [9] . We therefore , determined the subcellular localization of IRF3 following T . whipplei stimulation . Raw 264 . 7 macrophages overexpressing EGFP-IRF3 were incubated with T . whipplei for 4 h . Confocal microscopy allowed to visualized a marked nuclear translocation of IRF3 in response to T . whipplei ( Fig . 4A ) while , in unstimulated cells , IRF3 remained in the cytosol . This result suggests that the type I IFN induction depends on the transcription factor IRF3 . To further examine the role of IRF3 in type I IFN induction by T . whipplei , we used siRNA technology . Transfection of IRF3-specific siRNA ( si-IRF3 ) in Raw 264 . 7 macrophages resulted in a dramatic reduction of IRF3 levels at 24 h ( 84% ) , as determined by Western blot , while control scramble siRNA ( si-SCR ) had no effect ( Fig . S2A ) . IRF3-specific siRNA action was transient since IRF3 levels were back to normal 48 h post transfection . Therefore , we selected the 24 h time point to monitor the effect of IRF3 inhibition on IFN-β expression . Inhibition of IRF3 led to a profound reduction of IFN-β expression following T . whipplei stimulation , compared to control siRNA ( Fig . 4B ) . As a result , IFN-β production was reduced by 94% in cells lacking IRF3 ( Fig . S2B ) , thus identifying IRF3 as a component of the signalling pathway leading to type I IFN induction by T . whipplei-infected macrophages . In order to understand how T . whipplei turns on the type I IFN response , we examined the potential contribution of TLRs , the signalling of which is known to ultimately involve the adaptor molecules MyD88 and/or TRIF [25] . BMDM from double MyD88 and TRIF-deficient ( MyD88/TRIF−/− ) mice , which are unable to respond to TLR agonists , were stimulated with T . whipplei and IFN-β expression was monitored by qRT-PCR . Results showed that in contrast to wt BMDM , IFN-β expression was abrogated in MyD88/TRIF−/− BMDM ( Fig . 4C ) , suggesting that TLR signalling is required for type I IFN response during T . whipplei infection . Next , we wondered whether T . whipplei induces IFN-inducible genes via a type I IFN autocrine loop after engagement of the type I IFN receptor [26] . Thus , we first monitored the activation of the Stat1 transcription factor , one outcome of IFN secretion and type I IFN receptor engagement [27] . STAT1 activation was measured using specific antibodies targeting STAT1 phosphorylated at tyrosine 701 . As shown in Figure 4D , an increase in Stat1 Tyr701 phosphorylation after T . whipplei stimulation was evidenced at 6 h with further elevation at 24 h and 48 h . In contrast , Stat1 phosphorylation was completely absent when BMDM knocked-out for the type I IFN receptor gene ( IFNAR1−/− ) were stimulated with T . whipplei ( Fig . 4D ) . Subsequently , we analyzed the irg1 , ifit1 , ifit2 , ifit3 , mx1 and mx2 gene induction following T . whipplei stimulation in IFNAR1−/− BMDM . BMDM from wt and IFNAR1−/− mice were stimulated for 6 h and RNA were subjected to qRT-PCR . As expected , T . whipplei induced a marked expression of these genes ( Fig . 4E ) . However , the absence of type-I IFN receptor , which blocks the type I IFN autocrine induction , dramatically inhibited irg1 , ifit1 , ifit2 , ifit3 , mx1 and mx2 gene induction by T . whipplei ( Fig . 4E ) . Finally , only live bacteria induced type I response , as heat-killed forms of T . whipplei did not induce transcription of ifnb1 , irg1 , ifit1 , ifit2 , ifit3 , mx1 and mx2 ( Fig . 4F ) . Overall , these results indicate that type I IFN response is induced by viable T . whipplei organisms and likely involves a type I IFN autocrine loop . As members of the MAPK family are activated following the engagement of the type I IFN receptor and participate in the generation of IFN signals [28] , we treated BMDM with T . whipplei and MAPK activation was followed through their phosphorylation state at 24 h and 48 h . Interestingly , we found that p38 , ERK and JNK were phosphorylated at 24 h and their phosphorylation remained detectable 48 h after T . whipplei infection ( Fig . 5A ) . In order to examine whether this late MAPK induction was attributable to type I IFN signalling , we monitored the activation of p38 , ERK and JNK in IFNAR1−/− BMDM . After stimulation with T . whipplei , p38 and ERK activities were increased at 24 h and were still detectable at 48 h ( Fig . 5A ) . Conversely , the immunoreactive band of phospho-JNK was poorly if not detected in IFNAR1−/− BMDM stimulated for 24 h and 48 h ( Fig . 5A ) . Densitometry of the phosphorylated p38 , ERK and JNK 24-h autoradiographs confirmed the differences in band intensity ( Fig . 5B ) . These results suggest that T . whipplei induces a late MAPK signalling , which is , at least for JNK , dependent on the type I IFN receptor engagement . Type I IFNs are known to induce apoptosis , at least in part through up-regulation of tumor necrosis factor ( TNF ) family proteins such as CD95 ( Fas ) [29] . In addition , we previously showed that T . whipplei induces apoptosis of human macrophages and that circulating apoptotic markers are increased during active Whipple's disease [22] , [23] . To explore whether T . whipplei induces apoptosis of BMDM and whether type I IFN was involved , we measured BMDM apoptosis by TUNEL assay in time course experiments . We observed a gradual increase of TUNEL-positive cells that peaked at 18 h post infection , with nearly 25% of apoptotic cells ( Fig . 6A and 6B ) . Thereafter , the number of apoptotic cells slightly decreased and remained stable at 15% , 48 h post infection ( Fig . 6B ) . Interestingly , double-labeling of the apoptotic nuclei and T . whipplei in BMDM revealed that apoptosis was induced in infected cells , but also in cells that had not engulfed bacteria ( Fig . 6A , arrow ) . In IFNAR1−/− BMDM , T . whipplei-induced apoptosis at 18 h was significantly reduced as compared to wt BMDM ( Fig . 6C ) . This was not due to delayed apoptosis since incubating cells for longer periods did not reveal significant changes in cell death ( data not shown ) . In addition , UV exposure of IFNAR1−/− BMDM induced cell apoptosis at a level comparable to that of UV-treated wt BMDM ( Fig . 6C ) , ruling out the fact that the IFN receptor would have been required for apoptosis induction . Some studies have demonstrated that JNK plays a pivotal role in the activation of the apoptotic pathways [30] . As JNK was not activated and apoptosis was significantly reduced in T . whipplei-infected IFNAR1−/− BMDM ( see Fig . 5A and 6C ) , we wondered if JNK was required for T . whipplei-induced BMDM apoptosis . BMDM from wt mice were treated with the JNK specific inhibitor SP600125 for 30 min prior T . whipplei infection and apoptosis was measured after 18 h . We found that JNK inhibition significantly prevented T . whipplei-induced apoptosis ( Fig . 6D ) . Taken together , these results confirm that the transcriptional proapoptotic pattern induced by T . whipplei is functional and indicate that T . whipplei-induced apoptosis is dependent on an autocrine/paracrine loop involving type I IFN , its receptor IFNAR1 which leads to JNK activation . Puzzled by these findings , we wondered if bacterial replication was linked to type I IFN signaling . Thus , we infected BMDM from IFNAR1−/− mice with T . whipplei for 4 h and assessed bacterial replication . As expected , results showed that the type I IFN receptor was not involved in bacterial uptake , as around 7 , 000 bacterial DNA copies were detected after 4 h infection ( Fig . 7A ) , which were comparable to that found in wt BMDM ( Fig . 3A ) . However , replication of T . whipplei was reduced in these BMDM: around 10 , 000 bacterial DNA copies were detected at day 12 ( Fig . 7A , compare with Fig . 3A , 30 , 000 copies at day 12 ) , suggesting that type I IFN-dependent signaling is involved in macrophage permissivity to T . whipplei . Interestingly , we found that the killing of T . whipplei in IFNAR1−/− BMDM was associated with the maturation of T . whipplei-containing phagosomes , as T . whipplei colocalized with both Lamp1 and cathepsin D ( 85%±14% and 97%±5% , respectively ) at day 12 ( Fig . 7B ) . Finally , we wondered if type I IFN , JNK activation and bacterial replication were related . Hence , BMDM from wt mice were treated with a JNK-specific inhibitor . Bacterial survival and the nature of T . whipplei-containing phagosome were monitored by qPCR and confocal microscopy , respectively . JNK inhibition revealed cellular toxicity beginning at day 6 . However , during the first 6 days , bacterial replication was similar in untreated and SP600125-treated BMDM ( Fig . S3A ) . In addition , most bacteria colocalized with lamp1 ( 83%±13% ) , but not with cathepsin D ( 5%±3% ) at day 6 in both untreated and SP600125-treated BMDM ( Fig . S3B ) . Overall , these results showed that T . whipplei-induced type I IFN response governs bacterial replication through modulation of the phagosome conversion , independently of JNK activation .
A key requirement for dissecting the complex role of macrophages during infection is to understand how microbes activate or regulate host cells . In this study , we examined and characterized host responses induced by the facultative intracellular pathogen T . whipplei . Using microarray analysis of bone marrow-derived macrophages , we identified 59 genes that were significantly up-regulated upon infection . By bioinformatical approach , we found that most prominent GO groups covered immune response and cell communication . A closer analysis revealed that these over-represented genes could be classified in 4 functional categories . First , we found that T . whipplei induced M2 polarization of BMDM , which is consistent with the transcriptional profile of intestinal infiltrating cells , mainly comprised of macrophages , from patients with Whipple's disease [24] . Arginase , the M2 chemokines Ccl22 and Ccl17 , and the IL-1 receptor antagonist were induced in macrophage following infection as it has been described in Whipple's disease lesions [24] . M2 macrophages differ from classically activated M1 macrophages in terms of receptors , cytokine/chemokine expression , and effector functions . As a result , while M1 macrophages are microbicidal and inflammatory , M2 macrophages are rather seen as immunomodulators with diminished microbicidal activities [3] . We show that T . whipplei was able to invade and replicate within BMDM in a similar fashion to that observed in human macrophages [22] , suggesting that i ) mouse macrophages constitute model cells to study T . whipplei – macrophage interaction and ii ) T . whipplei-induced M2 polarization is a general response to this pathogen . Macrophage PRR responsible for T . whipplei recognition are still unknown . However , TLR2 and FPR2 , which encodes the mouse homolog formyl peptide receptor 2 of the human G-protein-coupled formyl peptide like receptor 1 were upregulated upon T . whipplei infection . TLR2 has been shown to be overexpressed in intestinal lesions of Whipple's disease [24] . Recently , TLR2 and the intracellular receptor nucleotide-binding oligomerization domain 2 ( Nod2 ) have been shown to cooperate in inducing the expression of FPR2 in microglial cells [31] . As FPR2 mediates the chemotactic activity of a variety of pathogen and host-derived peptides , it may actively participate in the macrophage infiltration observed in Whipple's disease lesions . Second and more strikingly , we found that a robust type I IFN response was induced by viable T . whipplei . As compared with the plethora of reports delineating the critical role of type I IFN in host resistance to many types of viruses , only few papers report their involvement during bacterial infections [32] . Results from our study suggest that type I IFN is induced in a MyD88-/TRIF-dependent pathway and demonstrate that , comparable to classical type I IFN triggering signals [33] , IRF3 signaling is activated following T . whipplei infection . Activation of the transcription factor IRF3 is likely to be dependent on TBK1 . Indeed , TBK1 is required for the activation and nuclear translocation of IRF3 in mouse embryonic fibroblasts ( MEF ) . Moreover , Tbk1−/− MEF show marked defects in type I IFNs , Cxcl10 , and RANTES gene expression after infection with either Sendai or Newcastle disease viruses or after engagement of the TLR3 and TLR4 by double-stranded RNA and LPS , respectively [34] . To our knowledge , type I IFNs have never been associated with the induction of M2 polarization of macrophages and are rather seen as M1 effectors [35] . However , our results strongly support this new association as data presented here suggest a positive feedback loop involved in BMDM response to T . whipplei . First , STAT1 was phosphorylated on Y701 , and this phosphorylation was absent when we used IFNAR1−/− BMDM . It has been shown that Stat1 activation is achieved by phosphorylation on Y701 that is followed by nuclear accumulation . For full transcriptional activity , Stat1 is also phosphorylated on S727 [36] . As T . whipplei induced the expression of several IFN inducible genes , it is likely that STAT was also phosphorylated on S727 . Second , we found that the transcription of these IFN inducible genes was abolished when macrophages lacking the type I IFN receptor were used . Our microarray analysis did not reveal any genes related to proinflammatory activities of macrophages , suggesting that T . whipplei is a weak inducer of inflammatory responses . Even if T . whipplei triggered a weak IκBα degradation , RelA translocation in the nucleus was not detected , despite strong translocation when cells were treated with LPS . This may be due to the lack of sensitivity of the immunofluorescence assay since IκBα , the gene of which is under the control of NF-κB is resynthetized and reached initial values by 3 h . Consistent with the weak activation properties of T . whipplei , we were not able to detect early MAPK signaling in macrophages . However , MAPKs were activated more lately , after 24 h . The kinetics of MAPKs activation suggest that p38 , ERK and JNK might be activated by a secondary signal , emanating from the initial T . whipplei-macrophage interaction . Indeed , type I IFN receptor engagement for example , has been shown to induce MAPKs [37] , [38] . Nevertheless , we found that only JNK phosphorylation was absent in IFNAR1−/− BMDM , while activation of p38 and ERK was similar to that observed in their wt counterparts . Another interesting feature of the T . whipplei - macrophage interaction revealed by this study is the induction of apoptosis . Macrophage apoptosis is probably linked to bacterial replication . Indeed , cells that are able to eliminate T . whipplei such as monocytes do not undergo apoptosis [22] . These results are strengthened by the fact that circulating levels of apoptotic markers such as nucleosomes are increased in patients with active Whipple's disease [23] . Here , T . whipplei induced BMDM apoptosis with a maximal response 18 h post infection , while heat-killed bacteria were unable to induce apoptosis ( data not shown ) . Induction of apoptosis appeared associated with i ) type I IFN response and ii ) JNK signaling . Apoptosis was inhibited by around 60% when IFNAR1−/− BMDM were infected with T . whipplei . In the meantime , JNK activation was abrogated in these cells . By using a JNK-specific inhibitor , we were also able to inhibit by 50% T . whipplei-induced apoptosis . Hence , we can hypothesize that T . whipplei induces type I IFN , which binds its receptor , induces JNK phosphorylation to promote macrophage apoptosis . Recently , Jeon and colleagues have shown that type I IFNs activate a JNK-specific signaling cascade involving Rac1 , MEKK1 , MKK4 and leading to apoptosis through filamin B [39] . Type I IFNs have also been shown to activate JNK for the induction of apoptosis in some lymphoma cells [40] . Finally , type I IFNs also activate the caspase cascade leading to apoptosis [41] . However , we cannot rule out the hypothesis that macrophage apoptosis arise from other signals . Indeed , we found that genes encoding Fas ( CD95/Apo1 ) and Tnfsf10 ( TNF-related apoptosis-inducing ligand , TRAIL/Apo2L ) were both significantly induced in BMDM in response to T . whipplei . Besides TNF itself , Fas and Tnfsf10 constitute two of the three death receptor/ligand systems that are responsible for the extrinsic induction of cell death [42] . Interestingly , Fas and Tnfsf10-dependent pathways involve JNK signaling and have been implicated in immunosuppressive and immunoregulatory functions [42] , [43] . Besides its role on apoptosis induction , type I IFN appeared to be involved in replication of T . whipplei . Bacterial replication was partly inhibited in IFNAR1−/− cells , as compared with wt BMDM . We also found that in wt BMDM , bacteria colocalized with Lamp1 but not with cathepsin D , as already described [44] . In contrast , in macrophage lacking the type I IFN receptor , bacteria mostly colocalized with cathepsin D . These results suggest that type I IFN can modulate , at least in part , microbial killing . Indeed , it has been shown that type I IFNs modulate vacuolar H+-ATPase-mediated acidification [45] . Interestingly , JNK activation was not required for T . whipplei replication and alteration of phagosome maturation . The role of JNK in phagosome conversion and bacterial killing is unclear as it seems to depend both on the upstream events ( engaged receptor ) and the pathogen itself . Indeed , it has been shown that JNK is involved in Staphylococcus aureus killing in a TLR2-dependent pathway through generation of superoxide , while its inhibition has no effect when cells are infected with E . coli [46] . From our study , two signals are emanating from the type I IFN receptor . The first involves JNK and leads to macrophage apoptosis while the second promotes alteration of phagosome maturation and bacterial replication independently of JNK . It has been shown that stimulation with type I IFN activates phosphatydilinositol-3 kinase ( PI3K ) and its downstream effectors [47] . As PI3K is involved in the modulation of phagosome maturation [48] , it is therefore possible that PI3K activity is modulated by T . whipplei to alter its phagosome and to favour its replication . Further studies are needed to determine from where these two signals diverge . A growing body of evidence shows that type I IFN participate in the host response to bacterial infection . However , their effects to the host can be either favorable or detrimental . For example , type I IFN response is critical in protecting the host against the extracellular pathogen group B Streptococcus [18] . In contrast , production of type I IFN during L . monocytogenes infection sensitizes macrophages to cell death [49] . Similarly , type I IFN production also appears detrimental for the host during infection with the T . whipplei-closely related M . tuberculosis [50] . M . bovis was shown to have enhanced replication rates in macrophages treated with type I IFN [51] . Our results suggest that the type I IFN induced by T . whipplei is detrimental for macrophages . Human infection with T . whipplei is a rare event despite the environmental ubiquity of the organism . Clinical features of Whipple's disease are non specific and it is clear that identifying the molecular mechanisms involved in type I IFN responses would have both clinical and therapeutic consequences .
BMDM from six week-old C57BL/6 and IFNAR1−/− [52] mice were isolated as described previously [53] . Double MyD88/TRIF-deficient mice were bred from MyD88−/− [54] and LPS2−/− [55] mice . Mouse RAW 264 . 7 macrophages ( American Type Culture Collection , ATCC N° TIB-71 ) were grown in Dulbecco's Modified Eagle Medium ( DMEM ) high glucose containing 10% FCS . The strain Twist-Marseille of T . whipplei ( CNCM I-2202 ) was cultured with HEL cells and purified as described previously [22] . Heat-killed T . whipplei was prepared by heating at 80°C for 1 h . All animal experiments followed the guiding principles of animal care and use defined by the Conseil Scientifique du Centre de Formation et de Recherche Experimental Médico-Chirurgical ( CFREMC ) and were approved by the ethics board of the university at which the experiments were performed ( Faculté de Médecine de la Timone ) . All experiments were performed at least three times . One representative experiment is shown . Error bars represent SD of triplicate values from a representative experiment . * , P<0 . 05 , Mann-Whitney's U test . The eGFP-IRF3 plasmid was kindly provided by G . Querat ( Marseille , France ) . RAW 264 . 7 macrophages were transfected with eGFP-IRF3 plasmid construct using Nucleofactor ( Amaxa Biosystems ) , according to the manufacturer's recommendations . IRF3-specific and control scramble siRNA were purchased from Santa Cruz Biotechnology . RAW 264 . 7 macrophages were transfected with IRF3-specific and control siRNA using Nucleofactor ( Amaxa Biosystems ) , according to the manufacturer's recommendations . T . whipplei organisms ( MOI 50∶1 ) were added to BMDM for 4 h , washed to remove free bacteria and incubated for 12 days in RPMI 1640 containing 10% FCS and 2 mM glutamine . Every 3 days , macrophages were collected and DNA was extracted using the QIAamp DNA MiniKit ( Qiagen ) . PCR was performed using the LightCycler-FastStart DNA Master SYBR Green system ( Roche ) , as previously described [22] . Macrophages seeded on glass coverslips were infected with T . whipplei ( MOI 50∶1 ) for 4 h , extensively washed to discard unbound bacteria and incubated in RPMI 1640 containing 10% FCS . At different time points , BMDM were fixed in 3% paraformaldehyde and permeabilized with 0 . 1% Triton X-100 . Immunofluorescence labeling was performed according to standard procedures [56] . Briefly , BMDM were incubated with rabbit anti-T . whipplei ( 1∶2 , 000 dilution ) antibodies ( Ab ) for 30 min [44] and rat anti-lamp1 ( 1∶1 , 000 dilution , clone 1D4B , purchased from DSHB ) or rabbit anti-cathepsin D ( 1∶1 , 000 dilution , a gift from S . Kornfeld , Washington University School of Medicine , St . Louis , Missouri ) . Secondary Alexa Abs were purchased from Invitrogen and used at a 1∶500 dilution . Coverslips were mounted with Mowiol and examined by laser scanning microscopy using a confocal microscope ( Leica TCS SP5 ) with a 63X/1 . 32-0 . 6 oil objective and an electronic Zoom 2X . Optical sections of fluorescent images were collected at 0 . 15-µm intervals using Leica Confocal Software and processed using Adobe Photoshop V7 . 0 . 1 . For the assessment of RelA ( p65 ) translocation in the nucleus , the same procedure was followed except that BMDM were incubated with rabbit anti-p65 ( RelA ) monoclonal Ab ( Cell Signaling ) . Cell culture supernatants were assayed for IFN-β by ELISA ( R&D Systems ) according to the manufacturer's instructions . BMDM were infected with T . whipplei for 6 h ( MOI 50∶1 ) and total RNA was extracted using the RNeasy minikit ( Qiagen ) . The quality and the quantity of RNA preparation were assessed using the 2100 Bioanalyzer ( Agilent Technologies ) . The 4X44k Mouse Whole Genome microarrays ( Agilent Technologies ) were used . Sample labeling and hybridization were performed according to the manufacturer recommendations ( One-Color Microarray-Based Gene Expression Analysis ) . Briefly , 300 ng of total RNA and cyanine 3-labeled CTP were used to synthesize labeled cRNA using the Low RNA Input Fluorescent Amplification Kit ( Agilent Technologies ) . Hybridizations were performed in triplicates for 17 h at 65°C using the In situ Hybridization Kit Plus ( Agilent Technologies ) . Slides were scanned at 5 µm resolution with a G2505B DNA microarray scanner ( Agilent Technologies ) . Image analysis and intra-array signal correction were performed using Agilent Feature Extractor Software 9 . 5 . 1 . 1 . Global normalization by trimmed means was applied on raw datasets using Excel ( Microsoft ) . Discrimination between samples was performed using the unpaired Student's t test . We only considered a gene as differentially expressed if the P value from Student's t test was below 0 . 01 and its absolute fold change was over 2 . To identify functional categories of genes that were over-represented in the data sets of modulated genes , we assigned Gene Ontology ( GO ) annotation by using the freely available online tools FatiGO Search ( http://babelomics . bioinfo . cipf . es/ ) and DAVID Bioinformatics Resources 2008 ( http://david . abcc . ncifcrf . gov/ ) . All transcriptional profile files have been submitted to the GEO database at NCBI ( accession number GSE16180 ) . cDNA was synthesized from 1 µg of total RNA using SuperScript II RNase H reverse transcriptase ( Invitrogen ) . Specific primers for each gene were designed using Primer3Plus , available online at http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi . The sequences of the targeted genes are listed in Table 2 . Quantitative RT-PCR was performed using LightCycler-Fastart DNA Master SYBR Green ( Roche Diagnostics ) and data acquired with the ABI PRISM 7900 HT ( Applied Biosystems ) . Gene expression was normalized to the β-actin gene , relative expression of respective genes was calculated using comparative threshold cycle method [57] . Macrophages were stimulated with either T . whipplei ( MOI 50∶1 ) or Escherichia coli 055∶B5 LPS ( 100 ng/ml , Sigma ) . At designated times , BMDM were washed with ice-cold PBS . Cells were then scrapped in ice-cold RIPA buffer ( 20 mM Tris-HCl , 200 mM NaCl , 1 mM EDTA , 1% Triton-X100 , pH 7 . 5 ) containing protease inhibitor ( Complete , Roche ) and phosphatase inhibitor ( Phosphostop , Roche ) cocktails . The cell lysates were cleared by centrifugation at 14 , 000 rpm for 15 min at 4°C and stored at −80°C . Cell lysates were examined for equal amounts of protein by the Bradford method using γ globulin as a standard [58] . Samples were loaded onto 10% sodium dodecyl sulfate polyacrylamide gels , electrophoresed and transferred onto nitrocellulose membranes ( Amersham ) . The membranes were blocked in PBS with 0 . 05% Tween 20 ( PBST ) supplemented with 3% powdered milk and then incubated with primary Abs against phospho-p38 , total p38 , phospho-ERK1/2 , total ERK1/2 , phospho-JNK , α-tubulin ( Cell signaling ) , IκBα ( Calbiochem ) or IRF3 Ab ( Santa Cruz ) as indicated by manufacturers . The blots were washed with PBST and incubated with a secondary Ab , either horseradish peroxidase-conjugated anti-rabbit or anti-mouse immunoglobulin ( Pierce ) in PBST plus 3% powdered milk . The bound Abs were detected using Immobilon Western Chemiluminescent HRP substrate ( Millipore ) . Detection of apoptosis by TUNEL was performed using In Situ Cell Death Detection Kit , TMR red ( Roche ) according to the manufacturer's instructions . JNK inhibition was performed using SP600125 ( Sigma ) at 50 µM for 30 min prior infection . As a control , apoptosis was induced by exposing cells to ultraviolet ( UV ) as described previously [59] . After treatment as indicated , cells on glass coverslips were fixed in 4% paraformaldehyde for 15 min , washed in PBS and permeabilized with 0 . 1% Triton-X100 in 0 . 1% sodium citrate for 2 min . Cells were then incubated with the TUNEL mixture containing TMR-dUTP and terminal deoxynucleotidyl transferase for 1 h . Cells were washed in PBS and nuclei were stained with DAPI before mounting with Mowiol . Positive controls were carried out by incubating cells with 3 U/ml DNase I prior labeling procedures . Negative controls were done by incubating cells with label solution ( without terminal deoxynucleotidyl transferase ) . Apoptosis was quantified as follows . Coverslips were examined in fluorescence mode with a Leica microscope equipped with a Nikon digital camera using a 10X objective lens . Three to five fields per condition ( 100 to 300 cells each ) were observed . The number of TUNEL-positive and DAPI-stained nuclei were determined and the apoptosis percentage was expressed as the ratio between TUNEL-positive and DAPI-stained nuclei ×100 . arg2 , 11847; ccl17 , 20295; ccl22 , 20299; clec4e , 56619; cxcl10 , 15945; fas , 14102; fpr2 , 14289; gbp2 , 14469; ifi44 , 99899; ifit1 , 15957; ifit2 , 15958; ifit3 , 15959; ifnar1 , 15975; ifnb1 , 15977; il1rn , 16181; irf3 , 54131; irg1 , 16365; mx1 , 17857; mx2 , 17858; myd88 , 17874; olr1 , 108078; tlr2 , 24088; tnfsf10 , 22035; trif ( ticam1 ) , 106759 .
|
Innate immune cells are sentinels allowing the host to sense invading pathogens . Among them , macrophages are highly microbicidal and are able to kill microorganisms . However , several pathogens have evolved strategies to hijack macrophage responses in order to survive or replicate . Tropheryma whipplei is the agent of Whipple's disease , a systemic disease that associates arthropathy , weight loss and gastrointestinal symptoms . It has been known for several years that this bacterium has a tropism for macrophages , in which it replicates . In this study , we have shown that T . whipplei induces host cell apoptosis and a surprising macrophage activation , characterized by anti-inflammatory molecules and type I interferon ( IFN ) signaling , which is generally associated to viral infections . We demonstrate that this type I IFN response is critical for bacterial pathogenicity , as it is required for bacterial replication and provides the first step of the apoptotic program of infected macrophages . By identifying these signaling events induced in macrophage by T . whipplei , we can now better understand the molecular basis of the pathophysiology of Whipple's disease , of interest for clinical and therapeutic ends .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"immunology/immunity",
"to",
"infections",
"microbiology/innate",
"immunity",
"immunology/leukocyte",
"activation"
] |
2010
|
Type I Interferon Induction Is Detrimental during Infection with the Whipple's Disease Bacterium, Tropheryma whipplei
|
It was not known how xeroderma pigmentosum group C ( XPC ) protein , the primary initiator of global nucleotide excision repair , achieves its outstanding substrate versatility . Here , we analyzed the molecular pathology of a unique Trp690Ser substitution , which is the only reported missense mutation in xeroderma patients mapping to the evolutionary conserved region of XPC protein . The function of this critical residue and neighboring conserved aromatics was tested by site-directed mutagenesis followed by screening for excision activity and DNA binding . This comparison demonstrated that Trp690 and Phe733 drive the preferential recruitment of XPC protein to repair substrates by mediating an exquisite affinity for single-stranded sites . Such a dual deployment of aromatic side chains is the distinctive feature of functional oligonucleotide/oligosaccharide-binding folds and , indeed , sequence homologies with replication protein A and breast cancer susceptibility 2 protein indicate that XPC displays a monomeric variant of this recurrent interaction motif . An aversion to associate with damaged oligonucleotides implies that XPC protein avoids direct contacts with base adducts . These results reveal for the first time , to our knowledge , an entirely inverted mechanism of substrate recognition that relies on the detection of single-stranded configurations in the undamaged complementary sequence of the double helix .
One of the most formidable challenges in DNA metabolism is that faced by the initiator of the nucleotide excision repair reaction as it locates damaged sites in the context of a large excess of mostly undamaged residues . This challenge is further complicated by an astounding diversity of target lesions , including cyclobutane pyrimidine dimers and pyrimidine–pyrimidone ( 6–4 ) photoproducts induced by UV ( ultraviolet ) light , bulky DNA adducts generated by electrophilic chemicals [1–4] , a subset of oxidative products [5–7] , and certain protein-DNA crosslinks [8] . Molecular defects in this versatile nucleotide excision repair response cause autosomal recessive disorders in humans such as xeroderma pigmentosum ( XP ) or Cockayne syndrome [9–11] . The XP syndrome , in particular , is characterized by photosensitivity and an extreme predisposition to sunlight-induced skin cancer [12] . In addition to cutaneous abnormalities , some XP patients also develop internal tumors [13] or neurologic complications leading to DeSanctis–Cacchione syndrome [14] . Individuals affected by XP are classified into seven repair-deficient complementation groups designated XP–A through XP–G [15] . The nucleotide excision repair response is separated in two pathways . Global genome repair ( GGR ) activity is responsible for the excision of DNA lesions across all nucleotide sequences , whereas transcription-coupled repair removes offending lesions only from the transcribed strand of active genes [16 , 17] . A principal difference between these pathways resides in the initial detection of DNA damage . During transcription-coupled repair , elongation of the RNA polymerase II complex is blocked by abnormal residues , thereby inducing the assembly of repair complexes [18] . In contrast , the GGR machinery is dependent on the initial recognition of damaged sites by XPC protein , which constitutes a universal sensor of bulky lesions [19 , 20] . Recent studies showed that XPC is also required for histone modifications in response to bulky lesion formation , presumably to facilitate chromatin remodeling [21 , 22] . It has been suggested that the recruitment of XPC protein is triggered by distortions of the DNA substrate [23–25] , but how this initial factor distinguishes between normal conformations of the double helix , induced by nucleosome assembly , transcription or other physiologic processes , and the DNA deformation at damaged sites remained elusive . This lack of mechanistic knowledge reflects the fact that no structure is available for any XPC homolog . Thus , the purpose of this study was to identify a nucleic acid interaction motif that is responsible for the unique recognition function of XPC protein . The human XPC gene encodes a polypeptide of 940 amino acids that exists as a complex with centrin 2 , a centrosomal protein , and HR23B , one of two mammalian homologs of yeast RAD23 . XPC protein itself possesses DNA-binding activity , whereas the centrin 2 and HR23B partners exert accessory functions [26 , 27] . Uchida et al . [28] have been able to narrow down the DNA-binding domain of XPC to a region of 137 amino acids ( codons 607–742 ) within its evolutionary conserved carboxy-terminal half . Because most mutated XPC alleles in xeroderma pigmentosum families lead to premature terminations as a result of frameshifts , nonsense mutations , deletions , insertions or aberrant splicing , only one single substitution , which causes a Trp690Ser change , has been identified in the evolutionary conserved region of XPC protein [29] . Although the loss of this aromatic side chain maps to the presumed DNA-binding domain , its consequence with respect to substrate recognition in the GGR pathway is unknown , prompting a mutational screen to analyze the general role of conserved XPC residues in the detection of DNA lesions . This study disclosed an aromatic hot spot , consisting of Trp690 and Phe733 , which mediates an affinity for the single-stranded character of target sites but with an astonishing aversion to associate with damaged DNA strands . A dual system of aromatics that stack with individual unpaired bases of single-stranded DNA has already been identified in RPA ( replication protein A ) , breast cancer susceptibility 2 protein , and many other single-stranded DNA-binding factors [30–33] . Therefore , our results point to a counterintuitive mechanism of damage recognition by which XPC protein avoids direct contacts with bulky lesions but , instead , probes the local susceptibility of intact nucleotides , on the opposite side of the double helix , to adopt a single-stranded configuration . The spontaneous Trp690Ser point mutation associated with the XP syndrome interferes with this inverted mode of substrate discrimination .
The human XPC sequence has been aligned [34] with its homologs from mouse , rat , Drosophila melanogaster , Trypanosoma cruzi , yeast , and Arabidopsis thaliana to identify potential consensus motifs in a region that includes the presumed DNA-binding domain [28] . This sequence alignment demonstrates that Trp690 , mutated in an XP family , is maintained from lower eukaryotes to plants and mammals . The only exception is provided by one of the two homologs in Schizosaccharomyces pombe , where the regular Trp at this position is replaced by another aromatic residue ( Figure 1 ) . The molecular function of an obligatory aromatic side chain at codon 690 was tested by a systematic comparison with all other evolutionary conserved aromatics that were identified in the same portion of human XPC protein , i . e . , between codons 531 and 742 . Also , the effects of these mutations were evaluated in relation to the substitution of other conserved residues with varying side chains . Figure 1 shows the positions in the presumed DNA-binding domain that have been selected for site-directed mutagenesis and highlights their degree of conservation among eukaryotes . A host cell reactivation assay was used to monitor the DNA repair proficiency of XPC mutants in human cells [35] . XP–C fibroblasts , which fail to express XPC protein , were transiently transfected with a dual luciferase reporter system accompanied by an expression vector coding for human XPC protein or the different mutants . The reporter construct , which carries a firefly luciferase gene , was damaged by exposure to UV light ( 254 nm; 1000 J/m2 ) and supplemented with an unirradiated control vector that expresses the Renilla luciferase . Following varying repair times , firefly luciferase activity was determined in cell lysates and normalized against the internal Renilla standard . Due to the repair defect of XP–C cells , transcription of the reporter gene was suppressed by persistent UV lesions , resulting in reduced firefly luciferase activity . However , DNA repair and , hence , firefly luciferase expression was restored following transfection with pcXPC , demonstrating that the genetic defect of XP–C fibroblasts is corrected by wild-type XPC protein ( Figure 2A ) . In contrast , expression of the reporter gene was not rescued when the same XP–C cells were transfected with the empty vector pcDNA ( Figure 2B ) . The residual background activity ( ∼15% of wild-type control ) , observed in the presence of these empty vectors , is likely due to the transcription-coupled repair process , which operates independently of XPC . In part , this residual activity may also result from a minor fraction of plasmids remaining free of bulky UV lesions in the luciferase reporter sequence . The firefly luciferase production was not restored when , instead of XPC , XPA protein was expressed in XP–C fibroblasts ( Figure 2B ) , thus demonstrating the specificity of our host cell reactivation system . Also , the firefly luciferase production was inhibited when the XPC sequence was modified to carry the Trp690Ser mutation responsible for clinical manifestations of the XP syndrome ( Figure 2B ) . Nearly identical results were obtained by transfecting the cells with vector pXPC–GFP , which drives the expression of wild type or mutated XPC sequences fused , on their carboxy-terminal side , to green fluorescent protein ( GFP ) . As expected , no complementation of the repair defect was detected upon expression of GFP alone using the corresponding control vector ( Figure 2B ) . The relative luciferase activity indicative of DNA repair was determined in the presence of each site-directed mutant , and the results were reported as the percentage of wild-type complementation after deduction of background luciferase expression . Initially , the aromatic side chains of conserved Phe , Trp , and Tyr residues were eliminated by Ala substitutions ( Figure 2C ) . In most cases , the excision-repair proficiency of XPC protein was only marginally diminished by these Phe→Ala , Trp→Ala , or Tyr→Ala changes . However , point mutations at the conserved codons 531 , 542 , 585 , 690 , and 733 resulted in a substantial ( >50% ) reduction of excision activity , and the residual DNA repair observed with these mutants is similar to the low level of complementation promoted by the Trp690Ser allele ( Figure 2C ) . All these mutants displayed essentially the same repair deficiency when reexamined as GFP fusion products ( unpublished data ) . The more sensitive codons 531 , 542 , 585 , 690 , and 733 were further tested by converting the respective aromatics to different amino acids with varying properties . In all cases , the luciferase activity reflecting DNA excision repair was strongly reduced regardless of whether the aromatics were replaced by the aliphatic side chain of Ala , the hydrogen moiety of Gly , or the hydrophilic side chain of Ser ( Figure 2D ) . These results imply that the loss of activity conferred by these XPC mutations is primarily a consequence of the missing aromatic residue rather than being dependent on the properties of the newly introduced substituent . Basic amino acids frequently make contacts with the phosphate moieties of the DNA backbone . Thus , evolutionary conserved Lys and Arg residues , located between codons 595 and 708 of the human XPC protein , were targeted by site-directed mutagenesis . The positively charged side chains were eliminated by changing the respective residues to Gly , but none of the resulting Lys→Gly or Arg→Gly substitutions were able to perturb the XPC function ( Figure 2E ) . In addition , absolutely conserved amino acids in the center of the putative DNA-binding domain of human XPC protein were changed to Ala residues . The resulting Pro635Ala , His644Ala , and Ser686Ala substitutions reduced the luciferase activity to a moderate degree but , interestingly , none of these mutants reached the low residual repair level observed after removal of an aromatic side chain at position 690 or 733 ( Figure 2F ) . The cellular XPC content was monitored by immunoblot analysis of XP–C fibroblasts harvested 15 h after transient transfections with vector pcXPC , promoting the expression of human XPC alone , or vector pXPC–EGFP translating to the production of XPC as a GFP fusion protein . In both cases , a quantitative comparison of protein levels demonstrated that the Trp690Ser and Trp690Gly mutants were expressed in human fibroblasts to similar levels as the wild-type counterpart ( Figure 3A and 3B ) . Moreover , the repair-deficient mutants with Ala substitutions at codons 531 , 542 , 585 , 690 , and 733 were detected in human fibroblasts in nearly identical amounts as wild-type XPC protein ( Figure 3C ) . Thus , the repair deficiency observed by substituting these conserved aromatics is not a consequence of reduced XPC expression or enhanced degradation . The GFP fusion partner was exploited to perform fluorescence microscopy studies . A time course experiment with the wild-type sequence demonstrated that expression of the XPC–GFP fusion increases during incubation periods of 18 h after transfection , with a cellular localization that is predominantly restricted to the nucleus ( Figure 3D ) . Control cells transfected with vector pGFP demonstrated that GFP alone displays a more diffuse distribution extending to both the cytoplasma and nucleus ( Figure 3E ) . However , the strong nuclear localization is reestablished after expression of GFP fused to the Trp690Ser mutant ( Figure 3F ) . A similar level of fluorescence with the same characteristic nuclear localization was recorded for each of the repair-defective Ala mutants ( Figure 3G ) . These results demonstrate that the repair deficiency of these tested mutants is not due to defective translocation into the nuclear compartment . The wild-type XPC polypeptide was coupled to maltose-binding protein ( MBP ) , produced in Spodoptera frugiperda ( Sf9 ) cells and purified to homogeneity by nickel and heparin affinity chromatography . MBP was chosen as a fusion partner to promote solubility and proper folding [36] . Another advantage of the MBP tag is that , on its own , it lacks DNA-binding activity [37] . On sodium dodecylsulfate gels , the final fraction of the MBP–XPC fusion product migrated as a single band with an apparent molecular weight of ∼170 kDa , which corresponds to the expected size of the 125-kDa XPC protein linked to the 43-kDa MBP moiety ( Figure 4A ) . Conflicting results regarding the affinity of XPC protein for DNA substrates of different lengths and conformations have emerged . Oligonucleotides with fewer than 60 base pairs resulted in weakened binding and reduced damage selectivity [25 , 38 , 39] . As a consequence , we employed radiolabeled duplexes of 65 base pairs to monitor DNA binding in electrophoretic mobility shift assays . The nucleotide sequence was designed to contain neighboring pyrimidines for the formation of UV-induced dimers . Thus , the double-stranded substrates were UV irradiated ( 254-nm wavelength ) to test the DNA damage selectivity of purified XPC fusion products . As expected from previous reports [40 , 41] , an increased affinity of XPC for UV-irradiated duplexes , over the unirradiated control , was detected when the binding reactions were supplemented with an excess of undamaged competitor DNA , i . e . , under conditions of limiting protein ( Figure 4B ) . In addition to this known affinity for UV-irradiated duplexes , we observed that XPC protein exhibits an extraordinary preference for binding to single-stranded 65-mer oligonucleotides over undamaged double-stranded fragments of the same length ( Figure 4C ) . These results obtained with relatively long oligomeric substrates imply that the XPC subunit fits the classic definition of a single-stranded DNA-binding protein . Because shorter duplexes are more prone to spontaneous denaturation , generating regions of single-stranded DNA , the preference of XPC protein for binding to single strands over double-stranded DNA is abrogated by reducing the oligonucleotide length to 40 residues or fewer ( unpublished data ) . This effect of substrate length provides a possible explanation for the diverging results of previous studies where the damage selectivity of XPC protein had not been attributed to an affinity for single-stranded DNA conformations [23 , 40] . A striking bias for single-stranded DNA is further supported by competition assays showing that the binding of XPC protein to UV-irradiated 65-mer duplexes is sensitive to the addition of 65-mer single strands ( Figure 4D ) . Conversely , when the competitor consisted of double-stranded plasmids , an excess of heavily UV-irradiated DNA was necessary to reduce the binding of XPC protein to single-stranded oligonucleotides ( Figure 4E ) . Subsequently , we observed that the high-affinity association of XPC protein with DNA single strands was progressively reduced upon UV irradiation of the oligonucleotide substrate ( Figure 4F ) . Interestingly , the UV dose of 600 J/m2 is expected to yield a damage frequency of <1 photoproduct/oligonucleotide molecule ( 40 ) , yet this low level of radiation was sufficient to reduce the single-stranded DNA-binding activity of XPC protein by ∼50% . Higher UV doses further suppressed the single-stranded DNA-binding activity to marginal levels ( Figure 4G ) , indicating that bulky lesions collide with the ability of XPC protein to form complexes with DNA oligonucleotides . Taken together , we conclude that XPC protein is recruited to target sites by virtue of its characteristic preference for deoxyribonucleotide sequences that adopt a single-stranded conformation . Surprisingly , this sensor protein associates preferentially with undamaged strands but rejects direct interactions with damaged strands . Two different strategies were used to test the ability of XPC mutants to interact with single-stranded DNA substrates . First , MBP–XPC fusion products were expressed in Sf9 cells , and the respective cell lysates were incubated with single-stranded DNA immobilized on agarose beads . After 2 h-incubations at 4 °C , the fraction of XPC protein in the pellet ( bound to DNA ) was separated by repeated washing from the free XPC molecules remaining in the supernatant . The extensively washed pellets and the accompanying supernatants were analyzed separately by gel electrophoresis and immunoblotting . Side-by-side comparisons showed that , in the case of the wild-type control , a major proportion ( >70% ) of XPC protein was recovered in the DNA-agarose pellet when the binding reactions were performed in buffer containing NaCl concentrations of 0 . 1–0 . 3 M ( Figure 5A , lanes 1–6 ) . If the NaCl concentration was raised to 0 . 4 M , only ∼50% of wild-type protein remained bound to DNA ( Figure 5A , lanes 7 and 8 ) . When the ionic strength was further increased , the proportion of XPC protein retained in the DNA pellet was diminished , reflecting a gradual reduction of nucleic acid binding . In the case of the Trp690Ser mutant , the fraction of protein recovered in association with the DNA beads was markedly reduced already in buffer containing 0 . 1 M NaCl ( Figure 5B , lanes 1 and 2 ) . When the NaCl concentration was increased to 0 . 2 or 0 . 3 M , the proportion of mutant XPC protein remaining in the DNA pellets was further reduced to ∼20% or less ( Figure 5B , lanes 3–6 ) . Essentially none of the Trp690Ser mutant remained assembled with DNA when the NaCl concentration was raised to 0 . 4 M ( Figure 5B , lanes 7 and 8 ) . These results show that the Trp690Ser substitution identified in an XP family disrupts the affinity of XPC protein for its DNA substrate . All repair-deficient substitutions were expressed as MBP fusion products and tested for their ability to interact with single-stranded DNA immobilized on agarose beads . This systematic comparison was performed in buffer containing 0 . 3 M NaCl , which corresponds to the ionic strength under which the most pronounced difference was detected between wild-type XPC protein and the Trp690Ser reference . Under these conditions , the three mutants Trp531Ala , Trp542Ala , and Tyr585Ala , which carry Ala substitutions outside the presumed DNA-binding domain , displayed a gradually reduced DNA-binding capacity compared to wild-type XPC protein ( Figure 6A ) , possibly reflecting indirect structural effects on the substrate recognition surface . This gradient of decreasing interactions with DNA culminated in the nearly complete loss of substrate binding in response to the Trp690Ala or Phe733Ala substitution . In both cases , the vast majority of mutant Trp690Ala and Phe733Ala protein appeared as free molecules in the supernatant , and only an insignificant fraction of these two species remained bound to the single-stranded DNA agarose beads ( Figure 6A ) . The Phe762Ala substitution , which yielded only a mild DNA repair defect in the host cell reactivation assay , was included in this nucleic acid-binding screen as an additional control . In full agreement with its in vivo repair proficiency , this Phe762Ala mutant was able to associate with the DNA substrate nearly as efficiently as the wild-type counterpart . Among the repair-deficient XPC mutants identified in this study , only the Phe733Ala substitution resulted in the same poor DNA-binding activity as the XP mutation at codon 690 . Therefore , an independent preparation of this Phe733Ala mutant ( Figure 6B , lanes 3 and 4 ) was reexamined for DNA binding in comparison with newly prepared cell lysates containing the repair-deficient Trp690Ser mutant ( lanes 1 and 2 ) , the repair-proficient Phe762Ala derivative , ( lanes 5 and 6 ) as well as the wild-type XPC control ( lanes 7 and 8 ) . This control experiment , again carried out in the presence of 0 . 3 M NaCl , confirmed that the removal of an aromatic side chain at positions 690 and 733 disrupts the DNA-binding function of XPC protein . Thus , the molecular defect underlying the prominent repair deficiency of these Trp690 and Phe733 substitutions resides with the inability of the respective mutants to undergo close contacts with the DNA substrate . A second experimental strategy , based on defined oligonucleotide probes , was established to confirm that the mutations at codons 690 and 733 confer defective binding to single-stranded DNA . For that purpose , MBP–XPC products were first purified from Sf9 cell lysates by immunoprecipitation with anti-MBP antibodies linked to paramagnetic beads . This one-step procedure generated nearly homogenous preparations of MBP–XPC fusion proteins ( Figure 7A ) . Subsequently , the amount of paramagnetic beads was adjusted to include 100 ng of purified protein , translating to a final XPC concentration of 3 nM in each binding reaction . Such purified fractions of wild-type protein or Trp690Ser mutant were incubated with radiolabeled 65-mer single strands and , following 2 h at 4 °C , the oligonucleotides captured by XPC protein were separated from free DNA . After extensive washing , the radioactivity associated with XPC protein on the paramagnetic beads was quantified by scintillation counting . We found substantial binding of wild-type XPC protein to single-stranded oligonucleotides but this interaction was markedly reduced when the Trp690Ser mutant was tested under exactly the same conditions ( Figure 7B ) . Next , the reaction mixtures were adjusted to contain different amounts of protein , thus demonstrating a dose-dependent increase of DNA-binding activity in the presence of wild-type XPC . These dose-dependence experiments confirmed that XPC protein interacts more efficiently with 65-mer heteroduplexes containing a 3-nucleotide bubble than to perfectly homoduplex controls ( Figure 7C ) . The DNA-binding activity was further enhanced by replacing duplex substrates with single-stranded oligonucleotides of the same length ( Figure 7C ) . Finally , these dose-dependent binding assays were used to compare the relative affinity of wild-type and mutant proteins for single-stranded DNA . In contrast to the efficient association of wild-type XPC with 65-mer oligonucleotides , the ability to interact with single-stranded DNA was essentially lost when we tested the mutants carrying an Ala substitution at codon 690 or 733 ( Figure 7D ) . However , in agreement with the different assay of Figure 6 , the DNA-binding activity was more moderately affected by a Trp531Ala substitution ( Figure 7D ) . These results support the conclusion that the two aromatic residues Trp690 and Phe733 are critically required for the recognition of single-stranded DNA conformations .
The most astounding feature of the GGR machinery is its ability to eliminate a wide diversity of DNA lesions , but how this repair system discriminates anomalous residues against the vast background of normal deoxyribonucleotides is still a focus of intense research , mainly because there is no common chemical motif among the different DNA adducts that would account for a classic “lock and key” recognition scheme [1–4] . Our mutagenesis screen designed to probe the mode of action of human XPC protein indicates that this primary initiator of the GGR reaction donates a pair of aromatic side chains ( Trp690 and Phe733 ) to monitor the double helical integrity of DNA and to recognize the local single-stranded character imposed on the undamaged side of the DNA duplex . These novel findings have several important implications with regard to damage recognition and the versatile GGR pathway . First , the preference of XPC protein for substrates containing a short single-stranded segment , over fully complementary duplexes , provides a truly universal mechanism for the detection of lesion sites . Normally , the native DNA duplex is stabilized by complementary base pairing as well as by stacking interactions between adjacent bases such that , in the absence of damage , the bases are positioned to the interior of the double helix . In contrast , DNA at damaged sites deviates considerably from this canonical Watson–Crick geometry . Bulky adducts often disrupt normal pairing and stacking interactions , thereby lowering the thermal and thermodynamic stability of the duplex , which results in local separation of the complementary strands and exposure of unpaired and unstacked bases on the surface of the double helix , thus generating an abnormal configuration with features that resemble single-stranded DNA [24] . The present equilibrium binding studies as well as kinetic measurements [25] , both demonstrating an extraordinary affinity for single-stranded oligonucleotides relative to double-stranded counterparts , imply that only base adducts that destabilize the double helix generate the key molecular signal for recognition by the single-stranded DNA-binding motif of XPC protein . Second , our results point to an inverted mode of recruitment mediated by an affinity for the undamaged strand of the DNA duplex . In fact , we observed an unfavorable binding of XPC protein to UV-irradiated DNA oligonucleotides compared to undamaged single-stranded counterparts . A similar reduction of oligonucleotide binding has been detected following the introduction of a site-specific cisplatin adduct [25] , implying that the interaction of XPC protein with single-stranded DNA is generally disturbed by the presence of adducted , crosslinked , or otherwise aberrant base residues . Thus , the exquisite affinity of XPC protein for single-stranded oligonucleotides , in combination with its aversion to interact with damaged strands , indicates that the recognition step in the GGR pathway is guided by an initial association with the native strand of damaged duplexes ( Figure 8A ) , without ruling out the possibility that XPC protein may ultimately interact with both strands . Such an inverted mode of damage recognition , which is completely independent of the variable chemistry of the lesion sites , accommodates the ability of the GGR machinery to detect a very wide array of DNA adducts . Recently , it has been reported that RPA is equally refractory to interactions with damaged oligonucleotides [42] , suggesting a functional analogy between XPC protein and representatives of the large family of single-stranded DNA-binding factors . Third , the dependence on a dual system of aromatic amino acids indicates a structural basis for the observed similarity between the XPC subunit and known single-stranded DNA-binding proteins . We found that two distinct aromatics in the presumed nucleic acid-binding domain of XPC protein , i . e . , Trp690 and Phe733 , are more critically involved in the high-affinity interaction with single-stranded configurations than all other conserved residues in the same XPC region . Even mutations affecting the absolutely conserved Pro635 , Lys642 , His644 , Tyr676 , Arg678 , Ser686 , or Lys708 , located in the DNA-binding domain , cause less incisive repair deficiencies than the removal of the aromatic side chains at positions 690 and 733 . Other aromatic side chains at codons 531 , 542 , and 585 are similarly required for excision repair activity , but their removal confers more moderate DNA-binding defects . This observation is consistent with a previous report indicating that residues 531–585 are located outside the core DNA-binding domain [28] . The distinctive requirement for a pair of aromatics ( Trp690 and Phe733 in the case of XPC ) is reminiscent of the OB-fold of many single-stranded DNA-binding proteins [30] . In RPA , for example , four different DNA-binding subdomains with the characteristic OB-fold are responsible for the association with single-stranded substrates [33] . Each of these domains forms a small β-barrel consisting of several short elements of secondary structure connected by loops of variable length [43] . The single-stranded DNA-binding activity of these RPA subdomains correlates with the presence of two structurally conserved aromatics that mediate stacking interactions with closely spaced DNA bases . Other OB-folds in the RPA complex that lack these aromatic side chains fail to contribute to nucleic acid binding [33] . The reiteration of a pattern of two separate aromatics in the DNA-binding domain of XPC protein lends support to the hypothesis that this repair factor may display an analogous structural fold to recognize DNA bases extruded from the double helix , and forced into a single-stranded conformation , as a consequence of bulky lesion formation . The different OB-fold subdomains of RPA range between 110 and 180 amino acids in length . As a minimal DNA-binding fragment of XPC protein has been mapped to a region of 136 amino acids [28] , we predict that XPC displays a monomeric variant of this motif to detect the single-stranded character resulting from separation of just one or , depending on the extent of DNA distortion , no more than a few base pairs at lesion sites . To summarize , XPC protein displays a range of properties that are typical of the OB-fold of single-stranded DNA-binding factors , i . e . , an affinity for single-stranded oligonucleotides , an exquisite preference for undamaged strands relative to damaged strands , the pairwise deployment of aromatics for nucleic acid binding , and the ability to interact with single-stranded DNA under conditions of elevated ionic strength . This combination of functional and structural analogies raises the question of whether a common sequence motif may be shared by XPC and known single-stranded DNA-binding proteins . A systematic analysis of the XPC full-length sequence did not reveal any signature that may have predicted its DNA-binding properties [44 , 45] . However , a homology search focused on the comparison with the growing family of OB-fold proteins showed that the nucleic acid-binding region of XPC protein displays a remarkable similarity to one of the oligonucleotide-binding subdomains of human RPA ( Figure 8B ) . This comparison yielded 27% identity and 73% similarity between the DNA-binding domain of XPC protein and the RPA-B motif situated in the large subunit of the human RPA complex . The sequence homology extends over most of the conserved elements of secondary structure of the RPA-B subdomain and exceeds the 12% identity detected when known OB-folds were aligned according to their high-resolution structure [32] . The same DNA-binding region of XPC also displays a 66% similarity with the OB1 and a 64% similarity with the OB2 motif of breast cancer susceptibility 2 ( unpublished data ) . Thus , the aromatic sensor domain of XPC protein , responsible for the recognition of DNA damage in the GGR pathway , is related to the OB-folds of known single-stranded DNA-binding proteins . In conclusion , this article shows that a versatile sensor of DNA damage achieves its wide recognition function by avoiding direct contacts with injured residues . Instead , XPC protein exploits the inherent redundancy of the genetic code in the DNA double helix to detect DNA damage in an indirect but highly versatile manner . If one strand contains a bulky lesion , normal base pairing and stacking interactions are compromised , and the intact complementary strand converts to a local single-stranded configuration , thus generating the universal molecular signal for XPC recruitment .
The human XPC complementary DNA [38] was cloned into pcDNA3 . 1 ( Invitrogen , http://www . invitrogen . com ) using the restriction enzymes NotI and KpnI and into pEGFP-N3 ( Clontech , http://www . clontech . com ) using the KpnI and XmaI sites . Mutagenesis was carried out with the QuickChange site-directed mutagenesis kit ( Stratagene , http://www . stratagene . com ) following the manufacturer's instructions . Forward and reverse primers are listed in Table S1 . The resulting clones were sequenced ( Microsynth , http://www . microsynth . ch ) to exclude accidental mutations introduced elsewhere in the complementary DNA . Simian virus 40-transformed human XP–C fibroblasts ( GM16093 ) were from the Coriell Cell Repository ( http://ccr . coriell . org ) . These cells were grown in Dulbecco's modified Eagle's medium ( Gibco , http://www . invitrogen . com ) , supplemented with 10% fetal bovine serum , penicillin G ( 100 units/ml ) and streptomycin ( 100 μg/ml ) , in a 5% CO2 humidified incubator . The pGL3 and phRL–TK vectors expressing firefly ( Photinus ) and Renilla luciferase , respectively , were from Promega . DNA was UV-irradiated at a concentration of 1 mg/ml in 10 mM Tris-HCl , ( pH 8 ) , and 1 mM EDTA . XP–C cells were transfected in a 6-well plate at a confluence of 95% using Lipofectamine Plus reagent ( Invitrogen ) . Each transfection mixture contained 0 . 23 μg pGL3 ( UV-irradiated ) , 0 . 02 μg phRL–TK ( unirradiated ) , and 0 . 25 μg of the appropriate expression vector . After a 4-h incubation , the transfection reagents were replaced by complete medium . Unless otherwise indicated , the cells were lysed after another 15-h period to measure firefly and Renilla luciferase activity using the Dual–Luciferase assay system ( Promega , http://www . promega . com ) on a microtiter plate luminometer ( Dynex , http://www . dynextechnologies . com ) . All results ( mean values of at least five determinations ) were normalized by calculating the ratios between firefly and Renilla luciferase activity . Expression of XPC polypeptides in human cells was monitored by Western blotting ( using monoclonal antibodies against GFP from Clontech ) and fluorescence microscopy as described [46] . A polyhistidine-MBP–XPC fusion product was constructed by inserting a 2 . 9-kb fragment , which contains the human XPC complementary DNA , into the pFastBac HTc vector ( Invitrogen ) using the NotI and KpnI restriction sites . Subsequently , a 1 . 2-kb fragment containing the MalE complementary DNA ( from pMal-c2; New England Biolabs , http://www . neb . com ) was inserted on the 5′ side of the XPC sequence using the StuI restriction site . Recombinant baculovirus for the infection of Sf9 cells was generated using the BAC-TO-BAC Baculovirus Expression System ( Invitrogen ) following the manufacturer's instructions . Polyhistidine- and MBP-tagged XPC protein was fractionated from Sf9 cell lysates [38] with two chromatographic cycles through a Ni2+ column ( Qiagen , http://www . qiagen . com ) . The fractions were analyzed using a mouse monoclonal antibody against recombinant human XPC protein ( Abcam , http://www . abcam . com ) . Samples containing XPC protein , eluting mainly at 100 mM imidazole , were pooled , dialyzed against phosphate buffer ( 25 mM sodium phosphate , [pH 7 . 8] , 10% [v/v] glycerol , 5 mM β-mercaptoethanol , and 0 . 25 mM phenylmethane sulfonyl fluoride ) containing 0 . 2 M NaCl , and further processed by heparin chromatography ( Amersham , http://www . amershambiosciences . com ) . The heparin column was eluted with a 0 . 2–1 M gradient of NaCl . The samples containing homogeneous MBP–XPC protein , eluting at 600 mM NaCl , were pooled , dialyzed , and supplemented with glycerol to a concentration of 25% ( v/v ) before freezing at −80 °C . Protein concentration was determined using the Bio-Rad protein assay reagent ( http://www . bio-rad . com ) . A one-step purification was performed by mixing crude Sf9 cell lysates ( 5–20 μl ) with monoclonal antibodies against MBP that were covalently linked to paramagnetic beads ( New England BioLabs ) . The binding buffer consisted of 25 mM Tris-HCl , ( pH 7 . 5 ) , 10% glycerol , 0 . 01% Triton X-100 , 0 . 25 mM phenylmethane sulfonyl fluoride , 1 mM EDTA , and 0 . 3 M NaCl . After incubation at 4 °C for 2 h , the beads were washed four times , and bound proteins were analyzed by denaturing gel electrophoresis followed by Coomassie staining . The yield of MBP–XPC protein was determined by quantitative laser densitometry of the 170-kDa bands using , as standards , different amounts of MBP–XPC probes purified by Ni2+ and heparin chromatography , as described before , and loaded in parallel onto the same gel . The synthetic 65-mer oligonucleotides 5′-CGGGGCGAATTCGAGCTCGCCCGGGATCCTCACATAGAGTCGACCTGCTGCAGCCCAAGCTTGGC-3′ and 5′-GCCAAGCTTGGGCTGCAGCAGGTCGACTCTATGTGAG GATCCCGGGCGAGCTCGAATTCGCCCCG-3′ were purchased from Microsynth . A DNA homoduplex was constructed by hybridizing these complementary oligonucleotides in 50 mM Tris-HCl , ( pH 7 . 4 ) , 10 mM MgCl2 , and 1 mM dithiothreitol . The annealing was performed by heating to 95 °C for 10 min , followed by slow cooling ( 3 h at 25 °C ) . Electrophoretic mobility shift assays ( reactions of 10 μl ) were performed by incubating , at 20 °C for 30 min , 32P-labeled oligonucleotide substrate ( 2 nM ) , duplex poly[dI–dC] competitor DNA ( 10 ng/μl ) , and the indicated concentrations of XPC protein in 40 mM Tris-HCl , ( pH 7 . 5 ) , 5 mM MgCl2 , 100 μg/ml bovine serum albumin , and 1 mM dithiothreitol [40] . Following the addition of gel loading buffer ( 2 μl ) containing 30% ( v/v ) glycerol , 0 . 25% ( w/v ) bromophenol blue , and 0 . 25% ( w/v ) xylene cyanol in water , the extent of binding was determined on 7% native polyacrylamide gels . Lysates ( 5 μl ) from baculovirus-infected Sf9 cells [28] were mixed with 50 μl of single-stranded DNA agarose beads ( Amersham ) and 100 μl 25 mM Tris-HCl , ( pH 7 . 5 ) , 10% glycerol , 0 . 01% Triton X-100 , 0 . 25 mM phenylmethane sulfonyl fluoride , 1 mM EDTA , supplemented by the indicated concentrations of NaCl . After incubation at 4 °C for 2 h , the supernatant was recovered and the beads were washed four times with 300 μl binding buffer . Finally , the DNA-bound proteins were eluted from the beads with 100 μl of 10 mM Tris-HCl , ( pH 8 . 0 ) , 1 mM EDTA , and 1% ( w/v ) sodium dodecylsulfate . Equivalent amounts of supernatant and DNA-bound fractions were loaded onto denaturing polyacrylamide gels , followed by immunoblot analysis , visualization by chemoluminescence ( SuperSignal , Pierce , http://www . piercenet . com ) , and quantification by laser scanning densitometry . The binding of mutants to single-stranded or double-stranded oligonucleotides was tested using purified MBP–XPC fusions obtained by immunoprecipitation . Paramagnetic beads ( 0 . 2 mg ) containing the indicated amounts of wild-type or mutant XPC ( between 10 and 100 ng ) were incubated with 32P-labeled 65-mer probes ( 2 nM ) in 200 μl of 25 mM Tris-HCl , ( pH 7 . 5 ) , 0 . 3 M NaCl , 10% glycerol , 0 . 01% Triton X-100 , 0 . 25 mM phenylmethane sulfonyl fluoride , and 1 mM EDTA . Following an incubation of 90 min at 4 °C , the paramagnetic beads were washed three times with 200-μl binding buffer . Finally , the radiolabeled oligonucleotides associated with XPC protein were quantified by liquid scintillation counting . The background radioactivity resulting from unspecific binding of the oligonucleotides to empty beads ( 0 . 2 mg ) was determined in separate reactions .
|
DNA is constantly exposed to damaging agents such as ultraviolet light , carcinogens , or reactive metabolic byproducts causing thousands of DNA lesions in a typical human cell every hour . To prevent irreversible mutations , many of these different lesions are eliminated by a DNA repair system known as “nucleotide excision repair . ” Repair is initiated by the XPC protein , which recognizes damaged sites in the DNA double helix . Here , we describe how the XPC protein probes the way in which the two DNA strands are aligned , and how a recurrent protein motif , termed oligonucleotide/oligosaccharide-binding fold , is used to detect dynamic fluctuations of DNA in the lesion containing regions . We show that XPC interacts preferentially with the undamaged strand opposite the lesion sites and conclude that XPC protein adopts an entirely indirect recognition mechanism to be able to detect a nearly infinite spectrum of DNA lesions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"eukaryotes",
"molecular",
"biology",
"mammals"
] |
2007
|
An Aromatic Sensor with Aversion to Damaged Strands Confers Versatility to DNA Repair
|
Antigen B ( EgAgB ) is a major protein produced by the metacestode cyst of Echinococcus granulosus , the causative agent of cystic hydatid disease . This protein has been shown to play an important role in modulating host immune responses , although its precise biological function still remains unknown . It is generally accepted that EgAgB is comprised of a gene family of five subfamilies which are highly polymorphic , but the actual number of genes present is unknown . Based on published sequences for the family , we designed specific primers for each subfamily and used PCR to amplify them from genomic DNA isolated from individual mature adult worms ( MAW ) taken from an experimentally infected dog in China and individual larval protoscoleces ( PSC ) excised from a single hydatid cyst taken from an Australian kangaroo . We then used real-time PCR to measure expression of each of the genes comprising the five EgAgB subfamilies in all life-cycle stages including the oncosphere ( ONC ) . Based on sequence alignment analysis , we found that the EgAgB gene family comprises at least ten unique genes . Each of the genes was identical in both larval and adult E . granulosus isolates collected from two geographical areas ( different continents ) . DNA alignment comparisons with EgAgB sequences deposited in GenBank databases showed that each gene in the gene family is highly conserved within E . granulosus , which contradicts previous studies claiming significant variation and polymorphism in EgAgB . Quantitative PCR analysis revealed that the genes were differentially expressed in different life-cycle stages of E . granulosus with EgAgB3 expressed predominantly in all stages . These findings are fundamental for determining the expression and the biological function of antigen B .
Antigen B ( EgAgB ) is the most abundant protein generated by the pathogenic larval stage ( hydatid cyst or metacestode ) of Echinococcus granulosus , the cause of cystic echinococcosis ( CE ) . Synthesized and secreted by both cyst germinal layer and protoscoleces [1] , the protein is highly immunogenic and can be recognised by more than 80% of sera from patients with CE [2] , [3] . Nevertheless , its precise biological function remains undetermined , although one report suggests that EgAgB might have lipid-binding properties [4] . It has been as well hypothesised that EgAgB plays a key role in the interaction between parasite and host based on studies showing it functions as a serine protease inhibitor that impairs neutrophil chemotaxis [5] and as an immune modulator that skews Th1/Th2 cytokine ratios to Th2 polarized responses [6] , benefiting parasite survival in the mammalian host [7] . A number of previous studies have also indicated that the protein is encoded by a gene family [8] , that is highly variable between isolates and strains of E . granulosus [5] , [8]–[10] . We believe the high levels of variation reported by others was based on comparisons of paralogs , amplified using conserved primers and assumed to be orthologs . Until now , there have been no data showing how many genes are represented in the EgAgB family , although it is known that there are five subfamilies ( EgAgB1-5 ) present [5] , [8] , [9] , [11]–[13] . Genomic Southern blots revealed that the gene family should include at least seven genes [14] . However , as these genes are highly similar , especially at the subfamily level , it has proven difficult to generate clear data from the Southern blot analysis . Determining the number of the genes in the family is fundamental for further exploring the expression and regulation of EgAgB in E . granulosus . This will provide insight to more fully understanding its biological function in this and other taeniid species , which share similar gene sequences to those found in E . granulosus [15]–[18] . We cloned and sequenced ten unique genes from individual worms ( adults and protoscoleces ) of E . granulosus and show that each is conserved in parasites originating from different geographical areas and hosts . Further , we show the differential expression of all of the family of genes in five developmental stages of E . granulosus by real time PCR and cDNA sequencing .
Protoscoleces ( PSC ) of E . granulosus were aspirated from a fertile hydatid cyst collected from a kangaroo ( Macropus giganteus ) from eastern Australia . The cyst was kindly provided by Dr . Tamsin Barnes from a previous study [19] . Mature adult worms ( MAW ) were collected from a dog from Xinjiang , China [20] . The parasite materials were stored until use in liquid nitrogen as described [20] . PSC and MAW were thawed in RNAlater ( Ambion , Austin , USA ) and diluted with water . Individual PSC and MAW were respectively pipetted into plastic mortar microtubes ( Sigma–Aldrich , St . Louis , USA ) under microscopy to make sure that each tube contained a single parasite . After a brief centrifugation to spin-down the parasite , 50 µl of PrepMan Ultra Sample Preparation Reagent ( Applied Biosystems , Foster , USA ) was added to each of the tubes . The single parasite was ground with a micro-grinder using a plastic pestle . The homogenate was heated at 100°C for 10 min and centrifuged at 16 , 000 g for 5 min . The supernatant was precipitated with 1× vol of isopropanol . The invisible pellet was washed with 70% ( v/v ) ethanol , dissolved in 50 µl water and used as DNA template . PCR reactions were performed with a Taq polymerase kit ( Promega , Madison , WI ) with 5 µl of the DNA template preparation and 20 pmol of each PCR primer in a final volume of 50 µl . To amplify the EgAgB gene fragments from genomic DNA , we designed two forward primers , EgAgBF1 , specific for subfamily EgAgB1 and EgAgB3 , and EgAgBF2 , specific for subfamily EgAgB2 , EgAgB4 and EgAgB5 based on previous studies [8] , [9] , [11] , [12] , [21] . The forward primers were based on the first exonic sequences of the EgAgB gene family . We designed eight down-stream primers , which were specific for each of the gene subfamilies ( the primers for EgAgB1 and EgAgB3 were within the second exons ) . All primers used to isolate the EgAgB gene variants are listed in Table 1 . Amplification was performed with 35 cycles of 94°C for 30 s , 54°C for 30 s and 72°C for 30 s , followed by a denaturing step at 94°C for 1 min , and a final extension step at 72°C for 7 min on a Mastercycle Gradient thermocycler ( Eppendorf , Hamburg , Germany ) . PCR products were purified using PCR Purification Kits ( Qiagen , Hilden , Germany ) . Fifty ng of the PCR products were ligated with 50 ng of pGEM-T vector ( Promega ) in a final volume of 20 µl according to the manufacturer's instructions . One microlitre of the ligation reaction was used to transform 20 µl of competent E . coli strain JM109 cells ( Promega ) . White colonies containing inserts were selected on LB agar plates containing ampicillin and X-gal . As each pair of primers is specific to each subfamily , and may amplify gene fragments with different sized PCR products , a quick plasmid extraction/PCR step was performed to determine the size of inserts before selecting clones for sequencing . In brief , after the white colonies had grown to about 0 . 5 mm in diameter , 30–50 of these colonies from each transformation were individually transferred to microtubes containing 50 µl of water . After vortexing and centrifugation at 12 , 000 g for 1 min , 10 µl of the supernatant from each tube was used as DNA template for PCR using the same original primers . For each transformation , 3–10 colonies with the same sized insert were selected for sequencing , performed using a Big-Dye Version 3 . 0 kit on an ABI 377 sequencer ( Applied Biosystems ) after purification with QIAprep Spin Miniprep Kits ( Qiagen ) . Inspection of the amino-acid sequences inferred from data collected during this study and obtained from the public databases showed that some members of the EgAgB subfamily could be aligned with ease . However , sequences from other subfamilies of EgAgB and sequences from other cestodes proved more difficult to align . Furthermore any alignment would be short: 54 amino acids being the length of the shortest sequence . However , to produce a graphical representation of the data , we constructed a simple phylogenetic tree to show the different clusters clearly , including relationships among members of each subfamily . We accept that such a tree does not provide robust inference for the deeper nodes . Bioedit ( http://www . mbio . ncsu . edu/BioEdit/bioedit . html ) was used to align sequences . Molecular Evolutionary Genetics Analysis version 4 ( MEGA v4 ) [22] program ( http://www . megasoftware . net/ ) was used to construct the tree from amino acid sequences translated from the second exonic sequences of EgAgB amplified and cloned from E . granulosus in this study and homologous protein sequences from other cestode parasites deposited in the GenBank , EMBL and DDBJ databases , after removal of the signal peptides at their N terminal . A distance matrix was constructed using a Poisson correction method before a mid-point rooted tree was constructed by the minimum-evolution method . One thousand bootstrap cycles were used . We used quantitative PCR to determine the expression level of each of the EgAgB family of genes in five life cycle stages/structural compartments of the cyst of E . granulosus . These were: protoscolex ( PSC ) , cyst germinal membrane ( CM ) , immature adult worm ( IAW ) , mature adult worm ( MAW ) and oncosphere ( ONC ) . Sheep livers containing hydatid cysts were collected from a slaughterhouse in Urumqi , Xinjiang , China . The inner parasite cyst membrane was carefully released from the outer host capsule under sterile conditions . PSC and brood capsules containing PSC were aspirated and then treated with 1% ( w/v ) pepsin in saline , pH 3 [23] , to remove capsule membranes and immature PSC . After 3 washes , the precipitated PSC were stored in liquid nitrogen until use . To prepare the CM , the inner cyst membrane was rinsed several times with PBS to remove any remaining PSC , and the membrane was divided into small pieces . These were pooled and stirred at 4°C for 30 min to release the germinal layer from the laminated layer . After leaving 1 min on ice , the laminated membranes were precipitated and the supernatant transferred to a fresh tube . After centrifugation at 3000 g at 4°C for 15 min , the pellet ( CM ) contained mostly germinal cells that were stored in liquid nitrogen until use . IAW and MAW ( from dogs infected with sheep PSC ) and activated ONC were prepared as described [20] , [24] . Total RNA was extracted from the different stages/compartments of E . granulosus using TRIzol reagent ( Invitrogen , Carlsbad , CA , USA ) , according to the supplier's instructions . The RNA was treated with DNase I ( Promega ) to remove possible genomic DNA contamination . All the RNA samples were of high quality ( A260/A280 nm>1 . 8 and <2 . 0 in nuclease-free water ) assessed using a Bioanalyzer RNA Pico LabChip ( Bioanalyer ) . First-strand cDNA synthesis was carried out with oligo ( dT ) 12–18 using a Superscript Reverse Transcription kit ( Qiagen ) with 45 ng of total RNA , according to the manufacturer's instructions . For real time PCR , all cDNA samples were diluted to a concentration of10 ng/µl . Subsequently , 5 µl aliquots were combined with 10 µl of SYBR Green , 3 µl of water and 2 µl ( 5 pmol ) of the forward and reverse primers listed in Table S1 . Each experiment was performed in triplicate . Expression profiles of EgAgB1-5 in the different stages/compartments were obtained by real time PCR using a Rotor Gene ( 6000 ) real time PCR machine ( Qiagen ) and data were analysed by Rotor Gene 6 Software . To identify the expression profile of EgAgB3 , we used a pair of primers , EgAgBF1 and EgAgB3R ( Table 1 ) , to amplify cDNAs obtained by reverse transcription from total RNA isolated from the five E . granulosus stages/compartments . The resulting PCR products were ligated into pGEM-T ( Promega ) and then transformed into E . coli strain JM109 . We randomly selected 30 colonies from each of the transformations for sequencing .
With the eight combinations of primers shown in Table 1 , we successfully amplified gene fragments with genomic DNA extracted from six individual MAW isolated from a dog ( from China ) and five individual PSC isolated from a single cyst from a kangaroo ( from Australia ) . Fig . 1 shows representative examples of the amplified bands from one MAW ( ZGA2 ) and one PSC ( ZGP5 ) . The sizes of the PCR products matched the predicted sizes ( 315 to 387 bp ) ( Table 1 ) . In total , we generated 435 clones with validated sequences including 234 from MAW and 201 from PSC . Alignment of all the sequences showed ten clusters ( data not shown ) representing ten genes . Figs . 2–4 show alignment s of intronic , exonic and amino acid sequences of ten gene representatives isolated from MAW ZGA2 and PSC ZGP5 , respectively . The terminology for each subfamily follows previous studies [5] , [8] , [9] , . Each pair specific to EgAgB1 , 2 and 5 generated only one sequence cluster , respectively , indicating only one gene in the three subfamilies , comprising subfamily 1 ( EgAgB1/1; accession numbers HM237302 ( PSC ) and GU166202 ( MAW ) ) , subfamily 2 ( EgAgB2/1; accession numbers GU166200 ( PSC ) and GU166201 ( MAW ) ) and subfamily 5 ( EgAgB5/1; accession numbers GU166215 ( PSC ) and GU166216 ) ( MAW ) ) . In contrast , primers specific to subfamily 3 amplified four genes in the subfamily ( EgAgB3/1–4 , accession numbers GU166204-GU166214 ) whilst primers specific for EgAgB 4 generated three genes in the subfamily ( EgAgB4/1–3 , accession numbers GU166196-GU166199 ) . Almost all the sequences in each gene cluster were identical to the EgAgB sequences deposited in GenBank ( Figs . 2–4 ) , which were obtained from isolates of E . granulosus from different geographical areas . Table 2 shows a comparison of each of the EgAgB DNA ( intron and second exon ) sequences and amino acid ( aa ) sequences encoded by the second exon , which are likely to be the mature and secreted proteins comprising 65–71 residues in length ( Table S2 ) . The degree of identity between the EgAgB protein family varies from 26 . 3% to 97 . 1%; the DNA sequences vary from 19% to 91% ( Table 2 ) . The lowest aa similarity occurred between EgAgB4/2 and EgAgB5/1 . Although EgAgB3/1 has the highest identity ( 97 . 1% ) with EgAgB3/2 at the aa level , the difference in their intronic sequences showed that they are different genes with a DNA identity of 57 . 1% . The major differences between the EgAgB genes appear in their introns ( Fig . 2 ) and the second exons ( Fig . 3 ) which encode different protein sequences ( Fig . 4 ) . The intronic sequences can be used for distinguishing all subfamilies and four genes in the EgAgB3 subfamily as they have different sizes and variable sequence ( Fig . 2 ) . Based on alignment analysis with sequences from GenBank , EgAgB1 has two clusters of intronic and exon sequences shown in Fig . S1 . They are likely to be encoded by different alleles . However , in our study , only one unique sequence ( EgAgB1/1 ) was amplified from individual worms and it has 89 bp of intronic sequence . The second exonic sequence comprises 198 bp ( Fig . 3 ) encoding 65 aa ( Table S2 and Fig . 4 ) . Cluster analysis of 99 cloned fragments of EgAgB2 with intronic and the second exonic sequences isolated from individual PSC and MAW ( 30 sequences aligned in Fig . S2 ) showed that the subfamily EgAgB2 comprises only one gene cluster , indicating there is only one gene in the EgAgB2 subfamily . The intron is 68 bp in length and the second exon is composed of 213 bp encoding 70 aa ( Table S2 ) . We designed a pair of primers to amplify the EgAgB3 gene subfamily by PCR from genomic DNA . Based on the size of inserts in clones and subsequent sequence analysis , we isolated four clusters of fragments representing four genes in the subfamily . EgAgB3/1 , EgAgB3/2 , EgAgb3/3 and EgAgB3/4 had introns of 137 bp , 140 bp , 152 bp and 140 bp respectively ( Fig . 2 and Table S2 ) . Although EgAgB3/2 had the same sized intron ( 140 bp ) as EgAgB3/4 , there were 26 substitutions between the two sequences . The second exonic sequences of the AgB3 subfamily also exhibited four types of sequences matching the intronic differences ( Figs . 2–4 ) . The amplified second exons of EgAgB3/1 and EgAgB3/3 encode 54 aa , but they are distinguishable from each other by differences in their intronic sequences of 137 and 152 bp , respectively ( Fig . 2 and Table S2 ) . In addition , there are eight aa substitutions in EgAgB3/1 compared with EgAgB3/3 ( Fig . 4 ) . The amplified regions of EgAgB3/2 and EgAgB3/4 encode 55 aa and 53 aa , respectively . The major difference in protein sequence encoded by the EgAgB3 subfamily occurs in the region immediately linked to the signal peptide , which is a region rich in aspartic acid ( D ) . EgAgB3/1 has 5Ds , EgAgB3/2 has 6Ds , EgAgB3/3 has 3Ds and EgAgB3/4 has 4Ds . Highly conserved sequences were found in the remainder of the second exonic sequences ( Figs . 3 , 4 ) . We designed four primers based on the 3′ terminal sequences of the EgAgB4 subfamily including one for the 3′ UTR sequence . Combined with forward primer , EgAgBF2 , the four pairs of primers allowed us to amplify three clusters of sequences from individual MAW , indicating that there are three genes ( EgAgB4/1–3 ) present in the subfamily . EgAgB4 is very similar to EgAgB2 both in intronic and exonic sequence ( Figs . 2 , 3 ) . The two subfamilies have the same sized 68 bp intron but there are ten nucleotide substitution differences in their intronic sequences ( Fig . 2 ) . In addition , there are 14–17 bp differences in the second exon of EgAgB4 compared with EgAgB2 ( Fig . 3 ) , resulting in 17 aa changes at the protein level ( Fig . 4 ) . The second exon of EgAg4/3 is composed of 216 bp encoding 71 aa , while the second exons of EgAgB 4/1 and EgAgB4/2 encode 70 aa and 69 aa , respectively ( Fig . 4 ) . The intronic sequences of EgAgB4 are identical . A major difference among the subfamily of genes is in their 3′ terminal exonic sequences , encoding different aa sequences rich in glutamic acid ( E ) and D residues ( Fig . 4 ) . EgAgB5 is a unique gene consisting of an intron of 67 bp and its second exon encodes a peptide of 66 aa ( Fig . 4 ) . Its DNA sequence is considerably different from those of the other EgAgB subfamily members ( Table 2 and Figs . 2 , 3 ) . Consequently , the protein sequence of EgAgB5/1 has the lowest identity to the other proteins ( Table 2 ) . We used MEGA methods for phylogenetic analysis of the inferred amino acid sequence of the EgAgB family of proteins to illustrate the evolutionary relationships within the family and , particularly , with those present in species from the confamilial genus Taenia . We confirmed the results with Bayesian analysis ( Mr Bayes 3 . 1 ) [25] ( data not shown ) and the two methods showed a very similar evolutionary pattern . The minimum evolution tree ( Fig . 5 ) has very low bootstrap values for deeper nodes , as anticipated because of dissimilarities between sequences from different subfamilies , and especially different species . The “Taeniidae antigens , ” [26] , commonly found in taeniid cestodes ( and one example from Hymenolepis diminuta ) form an outgroup in this mid-point rooted tree . All sequences from the genus Echinococcus , including sequences from E . granulosus ( EgAgB ) , E . multilocularis , E . vogeli , E . oligarthrus , E . ortleppi and E . canadensis form a monophyletic clade ( Fig . 5 ) . This implies that these genes have radiated in the Echinococcus lineage after separation from the other taeniids . For Echinococcus , the majority of the protein clusters include representative sequences from several species ( Fig . 5 ) , indicating the encoding genes were likely present in the most recent common ancestor of the genus suggesting the antigen B family has been important in its evolution . It is important to note that we treated all RNA preparations for analysis with RNase-free DNase prior to reverse transcription . To determine whether the RNA samples contained DNA after treatment , we added a PCR control that comprised the cDNA synthesis reaction comprising all components but without the addition of reverse transcriptase ( RT ) . Both normal RT and real-time PCR analysis showed there were no amplicons generated from these control samples ( data not shown ) . For normalizing the real time PCR data , we initially used actin II as a house-keeping gene to profile gene expression in the different stages of E . granulosus . However , as actin II was shown to be significantly up regulated in MAW and variable in the other stages , we used an eukaryotic translation initiation factor ( Eg-eif ) of E . granulosus as an alternative house-keeping gene , which was identified by microarray analysis and confirmed by real-time PCR and normal reverse transcription PCR analysis ( data not shown ) . Figure 6 shows the results of the expression levels of 5 subfamilies of the EgAgB genes and actin II after normalization using Eg-eif in the 5 E . granulosus stages and a pooled mixture of the 5 stages as a PCR control with different combinations of primers . EgAgB1 , EgAgB2 and EgAgB5 were expressed at very low levels in all stages . EgAgB3 was expressed in all stages of the parasite , with the highest in IAW and MAW . Except for EgAgB3 , the EgAgB genes were almost undetectable in PSC and ONC . EgAgB4 was expressed in CM , IAW and MAW , but at a low level . It is worth noting that EgAgB3 was highly expressed in MAW ( 3–10 times higher than in the other stages ) , suggesting this gene subfamily may play a role in worm development in the gut of the definitive host . We used EgAgBF1 ( Table 1 ) and EgAgB3R ( Table S1 ) sequences as universal primers to amplify cDNA which showed ( Table 3 ) that EgAgB3/1 was the most highly expressed gene in all stages , followed by EgAgB3/2 . EgAgB3/3 and EgAgB3/4 were lowly expressed .
All the genes in the E . granulosus antigen B ( EgAgB ) gene family have a similar gene structure with one intron flanked by two exons [27] . Furthermore , the first exonic sequence of EgAgB encodes a signal peptide . We did analysis of all EgAgB sequences deposited in the GenBank databases and showed that the sequences in this region are highly conserved ( data not shown ) with two clusters . This allowed us to design two forward primers , one for subfamily 1 and 3 , and another for subfamily 2 , 4 and 5 ( Table 1 ) in the first exonic region of the gene family . The variable sequences occur at the 3′ terminal ends . Consequently , we designed eight downstream primers specific to the 3′ terminal sequences to cover all possible genes in the five recognised gene subfamilies . Primer EgAgB24UTR ( Table 1 ) was designed based on the identical sequences of the 3′ terminal UTRs of subfamilies EgAgB2 and EgAgB4 , which allowed us to amplify the entire second exonic sequences in the subfamilies . With the designed primers , the PCR amplified fragments therefore contained both the intronic and the second exonic sequences of genes in the EgAgB family . Since eight pairs of primers were used to amplify genomic DNA from 11 MAW/PSC , instead of using random selection of clones for direct sequencing , we used a new strategy ( described in detail in the Methods and Materials section ) to select clones for sequencing . With this selection strategy , we chose 3–9 clones from each transformation for further sequencing . This strategy minimized the number of clones for sequencing and covered all possible EgAgB sequences . In total , we generated 435 clones with sequence information , which represents the largest reported number of EgAgB gene family sequences amplified from genomic DNA isolated from individual E . granulosus MAW and PSC . We isolated genomic DNA from individual PSC collected from a single hydatid cyst obtained from an infected kangaroo . The PSC clones allowed us to determine whether any apparent gene variation was caused by a different gene or by a mutation . As the PSC were collected from a single hydatid cyst , their genomic DNA should be identical [28] , and , indeed , we showed the sequences for each gene were indistinguishable . Two conclusions resulted from this sequence analysis: 1 ) . E . granulosus genomic DNA contains at least ten genes comprising the EgAgB family; and 2 ) . each of the genes is highly conserved . We isolated all ten genes from each of six MAW . The MAW were collected from a dog experimentally infected with pooled PSC originating from a number of hydatid cysts obtained from three individual sheep . The worms could , therefore represent different genotypes , but the sequence analysis showed that each gene was identical , confirming , therefore , the conservation of each gene in the EgAgB gene family , which was further supported by alignment with sequences deposited in the GenBank databases ( Fig . S3 ) . In addition , we showed that each of the ten EgAgB genes was identical in isolates collected from two distinct geographical areas , China and Australia . Macropods have only recently acquired E . granulosus as the parasite is believed to have been introduced into Australia by European immigrants about 200 years ago [29] . The conservation in sequence of the EgAgB genes isolated from a recently acquired new intermediate host , this case , a macropod , suggests that the EgAgB genes may play a fundamental role in parasite survival . EgAgB has been considered to be a polymorphic gene [5] , [8] , due likely to host selection for adaption given that E . granulsous strains are generally specific for the intermediate hosts they infect [28] . Accordingly , different stains have been presumed to have different genomic isoforms or alleles for some of their EgAgB genes [10] , [30] . An alignment with sequences from GenBank showed that EgAgB1 has two or three major clusters of intronic and second exonic sequences ( Fig . S1 ) . As the sequences have the same intronic and exonic sequence lengths and several nucleotide substitutions , they are likely to be encoded by a polymorphic gene that is strain-related [31] . It is not clear whether the variation of the sequence is due to heterozygosity , which has been shown in the Echinococcus malate dehydrogenase ( MDH ) gene [32] , or to the presence of host-specific alleles . We isolated one cluster of EgAgB1 sequence from the MAW and larval PSC of E . granulosus . The parasite samples were collected from different hosts from two continents . One sequence ( GU166203 ) was identical to one of the cluster sequences ( AF143813 cluster , Fig . S1 ) that is related to a sheep strain sequence [31] . Another two clusters in the EgAgB1 subfamily are related to those from E . granulosus cattle ( FJ696924-FJ696928 ) and buffalo ( FJ696936 , FJ696923 ) strains [30] , [31] . Further study is required to determine whether EgAgB1 can be used as a universal probe for distinguishing the recognized genotypes of E . granulosus [33] . It is not surprising that EgAgB comprises a multigene family . Southern blotting analysis showed several bands present in hybridizations with genomic DNA from E . granulosus [8] , [34] indicating the family has different genomic loci . With genomic DNA extracted from a single cyst , Chemale et al . [35] suggested there are three genes in the EgAgB gene family . Southern analysis , however , does not indicate precisely the number of genes in the family , which can only be determined by a sequencing approach . We performed a phylogenetic analysis of inferred amino acid sequence of EgAgB family proteins to illustrate the evolutionary relationships within the family and particularly with those of the confamilial genus Taenia spp . ( Fig . 5 ) . The Taenia proteins have been termed “Taeniidae antigens , ” as the encoding genes are found mostly in taeniid cestodes [26] , with one sequence ( AF249884 ) isolated from Hymenolepis diminuta , a member of the cyclophyllidean family Hymenolepididae . The proteins were classified into several major and distinct clusters . All sequences from the genus Echinococcus , including sequences from E . granulosus ( EgAgB ) , E . multilocularis , E . vogeli , E . oligarthrus , E . ortleppi and E . canadensis form a monophyletic clade ( Fig . 5 ) , which is separated from those of the large tape worms , such as Taenia and Hymenolepis . This suggests that these genes have radiated in the Echinococcus lineage after its separation from the other taeniids . This radiation might be correlated with the unique biological features of the Echinococcus genus such as the extensive asexual reproductive capacity of the multi-compartmentalized metacestode stage , the use of different hosts and organs for cystic development , small MAW with few segments and low definitive host specificity; perhaps some or all of these traits are indicative of a role for the antigen B proteins . Mumuti et al . [21] showed , using specific antibodies against each of 5 gene products in E . multilocularis , that the EmAgB genes were differentially expressed in the adult and larval cyst and PSC stages , with EmAgB3 being predominantly expressed; however , the ONC stage , which is responsible for human infection , was not included in the analysis . To determine the expression of the EgAgB family of genes in E . granulosus , we used quantitative PCR to measure their expression levels in the PSC , CM , IAW , MAW and ONC . As the genes in each of the subfamilies have very similar sequences , it was challenging to design PCR primers to readily distinguish them individually . However , the differences in sequences between the subfamilies allowed us to design specific primers to amplify cDNA fragments to distinguish the genes at the subfamily level . We initially used E . granulosus actin II ( accession no . L07773 ) as a house-keeping gene as used in other studies with Echinococcus [31] , [36]–[39] but this gene proved to be highly variable between different stages of the parasite at the transcription level , being expressed 35 and 20 times higher in MAW than in the ONC and CM , respectively ( Fig . 5 ) . Our results showed that the EgAgB gene family members were expressed differentially , with the EgAgB3 genes predominantly expressed in all life-cycle stages investigated , including the ONC . The expression profiles obtained were similar to these obtained by by Mamuti et al . [21] , for E . multilocularis , who used specific antibodies against the EmAgB protein family . We were able to demonstrate that there are 4 genes in the EgAgB3 subfamily . However , it is difficult to use normal real time PCR to distinguish their expression in E . granulosus due simply to the high similarity in their transcription levels . We expressed all the second exonic sequences of EgAgB3 and subsequent analysis showed that they cross reacted strongly ( data not shown ) , indicating neither normal real time PCR , nor Western blot analysis can be used for distinguishing each of the genes in the subfamily . Although not accurate , sequencing mRNAs from different stages of E . granulosus may be a way to predict the expression profiles of the EgAgB3 genes based on the transcription frequency of the genes . We demonstrated that EgAgB3/1 is the most predominant subfamily gene expressed in the intermediate host cyst and PSC stages , suggesting that EgAgB3/1 may be a suitable serodiagnostic target molecule . It is almost 40 years since the EgAgB protein was identified in E . granulosus hydatid cyst fluid [40] , but its precise biological function ( s ) still remains unknown . Here , we have shown that the E . granulosus antigen B family contains at least 10 genes . We believe these new findings are important for addressing the expression and regulation of the EgAgB genes , as they may provide new insights for determining the biological features and characteristics of the proteins encoded by this complex gene family , notably its potential role in the interaction between parasite and host as an immune modulator , benefiting parasite survival .
|
Antigen B ( EgAgB ) is a major protein produced by the metacestode cyst of Echinococcus granulosus and plays an important role in modulating host immune responses , although its precise biological function still remains unknown . Previous studies suggested the EgAgB gene family is variable between isolates and genotypic strains of E . granulosus . We designed specific primers to amplify and determine the number and variation of the genes using genomic DNA from individual worms . Based on sequence alignment analysis , we found that the gene family comprises ten unique genes . Each of the genes was identical in both larval and adult E . granulosus and in isolates collected from the two distinct geographical areas . We showed that the genes were differentially expressed in different stages of E . granulosus with one gene , EgAgB3/1 , expressed predominantly in all stages . This is the first study to report such a large number of unique and conserved genes in the EgAgB gene family and their differential expression in different life cycle stages of E . granulosus . These findings are fundamental for determining the expression and regulation of this gene family in E . granulosus and the biological function of antigen B .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"molecular",
"biology"
] |
2010
|
The Echinococcus granulosus Antigen B Gene Family Comprises at Least 10 Unique Genes in Five Subclasses Which Are Differentially Expressed
|
Osteoarthritis is one of the most frequent and disabling diseases of the elderly . Only few genetic variants have been identified for osteoarthritis , which is partly due to large phenotype heterogeneity . To reduce heterogeneity , we here examined cartilage thickness , one of the structural components of joint health . We conducted a genome-wide association study of minimal joint space width ( mJSW ) , a proxy for cartilage thickness , in a discovery set of 13 , 013 participants from five different cohorts and replication in 8 , 227 individuals from seven independent cohorts . We identified five genome-wide significant ( GWS , P≤5·0×10−8 ) SNPs annotated to four distinct loci . In addition , we found two additional loci that were significantly replicated , but results of combined meta-analysis fell just below the genome wide significance threshold . The four novel associated genetic loci were located in/near TGFA ( rs2862851 ) , PIK3R1 ( rs10471753 ) , SLBP/FGFR3 ( rs2236995 ) , and TREH/DDX6 ( rs496547 ) , while the other two ( DOT1L and SUPT3H/RUNX2 ) were previously identified . A systematic prioritization for underlying causal genes was performed using diverse lines of evidence . Exome sequencing data ( n = 2 , 050 individuals ) indicated that there were no rare exonic variants that could explain the identified associations . In addition , TGFA , FGFR3 and PIK3R1 were differentially expressed in OA cartilage lesions versus non-lesioned cartilage in the same individuals . In conclusion , we identified four novel loci ( TGFA , PIK3R1 , FGFR3 and TREH ) and confirmed two loci known to be associated with cartilage thickness . The identified associations were not caused by rare exonic variants . This is the first report linking TGFA to human OA , which may serve as a new target for future therapies .
In spite of advances in the understanding of OA , the absence of effective therapeutic targets demonstrates that a better comprehension of its causes and pathophysiological mechanisms is needed . Since genome-wide genetic studies are hypothesis-free and do not suffer from the bias of previous knowledge , they have the potential to identify novel biological pathways involved in OA . The discovery of novel genes has the potential to identify novel treatment options . In addition , more personalized medicine approaches for OA can be explored through prediction of risk for disease as well as classification of disease subtypes . Heritability of hip OA has been estimated to be around 40–60% . However , to date only few genetic variants have been successfully identified [1 , 2] . The reasons for finding only a modest number of genetic loci associated with hip OA can be attributed partially to relatively modest samples sizes in comparison to other complex diseases , such as myocardial infarction [3] . In addition , phenotype heterogeneity is an important issue in OA genetics as this is well known to reduce power to robustly detect signals . The problem of heterogeneity in genetic association studies of OA has been highlighted before [4 , 5] and is exemplified by the fact that the definition of the phenotype is a combination of bone and/or cartilage features as well as clinical complaints . Moreover , there is growing consensus that OA can be divided into multiple sub-phenotypes each with their own etiology and risk factors . For example , it has been demonstrated that individuals with hip OA , where only cartilage degradation is involved ( atrophic OA form ) , are linked to a different systemic bone phenotype compared to individuals with OA where bone formation is also present [6] . As a way to overcome this , we examined a quantitative trait , which is one of the structural components of joint health , cartilage thickness , as a distinct phenotype . Joint Space Width ( JSW ) is considered to be a proxy for cartilage thickness measured on hip radiographs . Minimal JSW ( mJSW ) has been shown to be a more reliable measure for hip joint health compared to the classical Kellgren & Lawrence score [7] . Previously , we have demonstrated that using only a modest discovery sample size ( n = 6 , 000 ) , we were able to successfully identify a genome-wide significant association of the DOT1L locus with mJSW as well as hip OA [1 , 8] . We now aimed to perform a more powerful analysis by combining data from five studies in the discovery phase , and subsequent replication in seven additional studies , amounting to a total sample size of 21 , 240 to identify new genes implicated in joint health using mJSW as a proxy for cartilage thickness . Using whole exome sequence data from 2 , 050 individuals we screened the discovered genes for potential functional variants . Subsequently we used multiple approaches that leverage different levels of information to enforce evidence of candidate genes annotated close to the associated signals .
Genome-wide Association analysis of mJSW of the hip with genetic variants was performed in a discovery set that included 13 , 013 individuals ( see S1 Table and S1 Text for cohort specifics ) with data on ±2 , 5 million genotyped or HapMap Phase II imputed SNPs . We applied extensive quality control measures ( see S2 Table and S3 Table for details on quality control and exclusion criteria ) leaving a total of 2 , 385 , 183 SNPs available for association analyses . Genomic control inflation factors for the P values of the RS , TwinsUK , MrOS , and SOF GWAS were low ( λ = 1 . 02 , 1 . 01 , 1 . 02 and 0 . 99 respectively ) , and the interquantile-quantile plot ( S1 Fig ) also indicated no residual population stratification due to cryptic relatedness , population substructure or other biases . The discovery analysis yielded eighteen independent SNPs with suggestive evidence for association ( P <1*10−5 ) with mJSW , of which five ( four genetic loci ) met the genome-wide significance threshold of P-value ≤ 5*10−8 ( see Fig 1 ) . The top SNPs from these eighteen loci were selected for replication in additional 8 , 227 individuals from seven different cohorts . We observed that six of the eighteen SNPs significantly replicated ( P<0 . 05 ) with the same direction of effect ( see Table 1 ) . When we combined discovery and replication results in a meta-analysis , the five SNPs that met genome-wide significance in the discovery analysis became more significant and another two SNPs that replicated in independent cohorts reached suggestive evidence ( P≤ 1*10−6 ) for association in the combined meta-analysis . The top signal in the combined meta-analysis , rs1180992 ( Table 1 , Pcombined = 3 . 2x10-16 ) , is located in the intronic region of the previously OA associated DOT1L gene . This variant is very close to and in linkage disequilibrium with rs12982744 ( D’ = 1 , r2 = 1 ) , which was previously found in association with mJSW and hip OA [1 , 8] . The DOT1L signal was followed in strength of association by rs2862851 ( Pcombined = 5 . 2x10−11 ) , which is annotated to the intronic region of TGFA ( Fig 2A ) . Two variants near RUNX2 , rs10948155 and rs12206662 , also reached genome-wide significance for association with mJSW ( Fig 2B ) . The two variants in the RUNX2 locus were weakly correlated ( r2<0 . 2 ) . Conditional analysis , using GCTA , showed that both SNPs represented different signals ( S4 Table ) . Finally , the last signal that reached genome-wide significance was rs10471753 , an intergenic variant closer to PIK3R1 ( ~450 Kb ) than to SLC30A5 ( ~750Kb ) ( Table 1 , Pcombined = 3 . 8*10−9 ) . Other suggestive signals for association with mJSW at a Pcombined≤ 1x10−6 including signals with significant replication were rs496547 ( p = 1 . 5x10−7 ) , a downstream gene variant located 3' of TREH and , an intron variant annotated near SLBP ( rs2236995; p = 9x10−7 ) . All other additional signals selected in the discovery stage did not replicate . We examined whether the five GWS and two suggestive mJSW loci were also associated with hip OA in a total of 8 , 649 cases and >57 , 000 controls . Detailed description of the cohorts and OA definitions is given in S1 Table . Table 2 shows the associations found with hip OA . We observed that five of the seven identified mJSW loci were also associated with hip OA ( p-value<0 . 05 ) . Apart from the known DOT1L locus , the variant near TGFA was significantly associated with hip OA ( Table 2 , P = 4 . 3x10−5 ) . In addition , the SNP near SLBP and the two SNPs near RUNX2 were associated with hip OA . One of the latter SNPs , rs10948155 , is in high LD with a variant ( rs10948172 , D’ = 0 . 95 and r2 = 0 . 90 ) previously found in association with hip OA in males at borderline GWS level ( 2 ) . However , in our study , rs10948155 was just marginally associated with hip OA in the overall analysis ( Table 2 , P = 0 . 021 ) . We observed the second variant in this genomic region , an intronic variant in RUNX2 , rs12206662 , to have a larger effect size ( β = 0 . 14 , P = 1 . 1×10−4 r2 = 0 . 09 with rs10948172 ) . We further examined whether the identified loci were found associated with other phenotypes in earlier reports ( Table 3 ) . Five of the seven identified mJSW SNPs mapped to loci that have previously been associated with other bone-related phenotypes , primarily height . However , many of the identified height loci were not highly correlated with the mJSW signal ( Table 3 ) . Additional adjustment for height did not have an effect on the described association with mJSW; they showed an independent , possibly pleiotropic effect , on both traits . A particularly dense number of associations with different bone related phenotypes were present in the RUNX2 5’ region , where variants have been associated to BMD [10] , height [11] , osteoarthritis [2] and ossification of the spine [12] . Given the low LD between the variants underlying the different GWAS signals , it is likely that these represent independent associations . We used multiple approaches that leverage different levels of information ( e . g . , gene expression , regulatory regions , published literature , mouse phenotypes ) to prioritize candidate genes at each mJSW locus . Table 4 shows the summarized results from these analyses . In addition to the seven loci identified in the current study , we also analyzed five previously published loci for hip OA [2] . First , we focused on two gene prioritization methods: ( i ) DEPICT , a novel tool designed to identify the most likely causal gene in a given locus , and identify gene sets that are enriched in the genetic associations [21] , and ( ii ) GRAIL which uses existing literature to identify connections between genes in the associated loci [22] . The DEPICT analysis yielded seventeen significantly prioritized genes ( FDR >0 . 05 ) , from which three genes were also significantly prioritized in the GRAIL analysis ( S5 Table and S6 Table ) . Next , using the Online Mendelian Inheritance in Man ( OMIM ) database ( http://omim . org ) , we identified genes with mutations implicated in abnormal skeletal growth in humans; for 50% of the loci , a skeletal syndrome gene was present ( S7 Table ) . Similarly , we investigated if any of the genes had a known bone and cartilage development phenotype in mice . Very similar to the human phenotypes , the mice knockouts of the same genes resulted in bone and cartilage phenotypes ( http://mousemutant . jax . org/ ) ( S7 Table ) . Other supporting biological evidence that we gathered consisted of known expression quantitative loci ( eQTL ) and nonsynonymous variants in LD ( r2>0 . 6 ) with the lead SNP of a locus ( S8 Table and S9 Table ) , as well as expression in bone and cartilage tissue in mice using data from the Jackson lab database ( S7 Table ) . To further explore which genes are possibly underlying the genetic associations identified in this study , we analyzed gene expression in a paired set of non-lesioned and OA-lesioned cartilage samples of the RAAK study acquired from 33 donors at the time of joint replacement surgery for primary OA [19] . We first examined which genes are expressed in a set of seven human healthy cartilage samples ( S7 Table ) . Additionally , we tested which of the genes located in 1MB region surrounding the lead SNP were differentially expressed in OA-lesioned cartilage versus non-lesioned cartilage of the same hip . Of the 152 genes that were selected , 129 genes were represented on the array . Of those , 64 genes were significantly expressed in the cartilage samples . For eight of the twelve loci , we found genes that were differentially expressed in OA lesioned cartilage versus non-lesioned cartilage ( Table 4 , S10 Table ) . Differential expression in cartilage healthy vs OA affected cartilage was performed likewise ( S10 Table ) , while additionally adjusting for sex and age . Given the relatively small number of healthy samples ( n = 7 ) with large age range these data are less robust and we did not use these data in gene prioritization . For each gene a prioritization score was computed , based on equally weighting of the ten lines of evidence ( Table 4 ) . Following this approach , RUNX2 is highly likely to be the causal gene associated with rs12206662 and rs10948155 . Similar strong evidence is found for rs788748 ( IGFBP3 ) and rs10492367 ( PTHLH ) . In addition , suggestive evidence for a causal gene is found for the following: rs10471753 ( PIK3R1 ) , rs835487 ( CHST11 ) , rs2862851 ( TGFA ) , rs6094710 ( SULF2 ) , rs9350591 ( COL12A1 ) and rs11177 ( GNL3 ) . However for some loci the current evidence is ambiguous , suggesting more than one gene as the potentially causal one; rs2236995 ( FGFR3 or SLBP ) , rs11880992 ( GADD45B or DOT1L ) and rs496547 ( KMT2A or UPK2 ) ( Table 4 ) . In 2 , 628 individuals from the Rotterdam Study , exome sequencing was performed at a mean depth of 55x . Of those , 2 , 050 individuals also had mJSW and hip OA phenotype data . Baseline characteristics of those individuals were similar to the source population , mean age was 67 . 3 years , 57% of the individuals were female and mean of mJSW was 3 . 81 mm ( SD 0 . 82 ) . Details of the experimental procedure and variant calling are given in the supplementary material ( S2 Text ) . Only the variants with a minimal allele count of three in the total population were selected for analysis . Within the sixteen prioritized genes , a total number of 158 variants were identified in the protein-coding region , of which 85 were non-synonymous and one was a stop-gain mutation ( Table 5 , S11 Table ) . We first performed a single variant test , where we tested each of the 86 variants changing the amino-acid sequence for association with the mJSW trait ( S11 Table ) . We observed four nominal significant associations , with rare variants in SULF2 , TGFA , RUNX2 and FGFR3 . None of these rare exonic variants explained the original association between the GWAS hit and mJSW or hipOA when tested in a multivariate model ( S12 Table ) . Next , we performed a burden test ( SKAT ) [23] , to investigate whether the cumulative effects of the variants present in the sixteen selected genes were associated to mJSW , while adjusting for age and gender ( Table 5 ) . We observed a nominal significance burden test ( p<0 . 05 ) for TGFA , SULF2 , CHST11 and RUNX2 for mJSW . However , none of these findings reached significance after correction for multiple testing . For most of the loci , no obvious protein-coding variants were found that could explain the associations . In previous studies it was shown that disease-associated variants are enriched in regulatory DNA regions [24 , 25] . We therefore examined whether the identified DNA variants ( or SNPs in high LD ) resided in chondrocyte and/or osteoblast specific enhancer regions , using data from ENCODE and ROADMAP [26–28] . To this end , we compared CHIP-seq signals from five different chromatin state markers ( H3K4me3 , H3K4me1 , H3K36me3 , H3K27me3 , H3K9me3 ) in chondroblasts and osteoblasts to four cell lines from another origin . Together , these chromatin state markers identify promoter and enhancer activity in each of the cell lines . With the exception of rs2862851 , we observed that for all mJSW genetic loci , SNPs in high LD were located in cell regulatory regions in chondroblast and/or osteoblast cells ( see Fig 3 for an overview and S13–S18 Tables for each locus ) .
Only a modest number of genetic variants has been successfully identified through genome-wide association studies for OA This can in part be explained by the phenotypic heterogeneity of OA . Therefore , we used mJSW , a proxy for cartilage thickness in the hip joint , as one of the structural components of joint health . An additional advantage of this phenotype is its continuous nature , which increases power compared to a dichotomous trait , such as OA-status . We identified six independent loci associated with cartilage thickness in the hip joint , of which four surpassed genome-wide significance ( TGFA , PIK3R1 , SUPT3H-RUNX2 , DOT1L ) and two were suggestive for association with mJSW ( SLBP/FGFR3 , TREH-DDX6 ) . Four of these loci ( TGFA , SUPT3H-RUNX2 , DOT1L and FGFR3 ) were also associated with hip OA . The fact that we were able to identify six loci with the current sample size ( 13K individuals in the discovery ) indicates that cartilage thickness is a phenotype providing a better yield in number of discoveries than the efforts ran with traditional composite radiographic scores . As a comparison , the largest GWAS study up to now , arcOGEN with 7 , 4K cases and 11K controls as discovery , yielded one locus in the overall analysis , and seven additionally in a number of stratified analyses . Interestingly , in the current manuscript we report on rs10948155 , which is in high LD ( r2 >0 . 8 ) with a locus from arcOGEN which was only marginally associated ( p below genome-wide significance threshold ) with OA in males only [2] . By using a cartilage specific endophenotype , evidence for this locus is elevated here to genome-wide significance in the total population , underscoring the increased power when more specific endophenotypes are used . Endophenotypes are quantifiable biological traits intermediate in the causal chain between genes and disease manifestation ( in this case osteoarthritis ) . JSW can be precisely measured throughout the life of individuals [7] and also displays variation in normal subjects . Therefore , mJSW may be more tractable for the genetic dissection of OA . Across the cohorts in this manuscript , mJSW has been measured in different ways , using both hand measured JSW on radiographs as well as ( semi ) automatic software which could have added some noise to the overall meta-analysis . Future cross-calibration of JSW measurements might aid in a more precise measurement and additional power to pick up genetic loci . To the best of our knowledge , we are the first to scrutinize exome variants in relation to OA identified by large scale re-sequencing . We did not find low frequency exonic variants in any of the prioritized genes that could explain the observed associations with mJSW . We do have to keep in mind that the power of the exome sequencing effort is smaller than the original discovery analysis . We were unable to examine variants with allele frequencies below 0 , 07% . In addition , for rare or low allele frequencies , we only had power to detect relatively large effect sizes . For example , we had 80% power to detect a beta of 0 , 7 mm ( almost 1SD ) difference for a variant with 1% allele frequency . However , we tested all of the discovered exome variants in a multivariate analysis , and found that the novel identified rare exome variants did not affect the association between the GWAS-identified variants and mJSW in the same sample . This suggests that the associations between mJSW and the identified SNPs are not explained by rare exonic variants and likely exert their effects through regulation of expression . Indeed , supporting this hypothesis , we found that many these variants ( or SNPs in LD ) were annotated in regions that were annotated as regulatory active in chondroblastic and/or osteoblastic cells . However , more work is needed to examine the exact biological mechanism underlying the identified genetic loci . TGFA ( Transforming Growth Factor Alpha , rs2862851 ) was the strongest novel locus associated with cartilage thickness and hip OA . TGFA has been suggested to be involved in endochondral bone formation in mice , specifically the transition from hypertrophic cartilage to bone [29] . Recent , TGFA has also been implicated in the degeneration of articular cartilage during OA in rats [30] . Our results now imply a relationship between TGFA and human OA . In addition to the genetic association , we also show that TGFA expression is higher in human OA affected versus non-lesioned cartilage , possibly indicating that TGFA has a role in cartilage remodeling . Functional characterization of the TGFA- associated locus by an examination of the histone methylation marks representing promoter or enhancer activity , did not reveal an obvious explanation for the functional impact of the SNP . However , the examined histone mark data represent unstimulated cells , and it is anticipated that the promoter and enhancer activity change upon stimulation of the cells . It is becoming more clear that effects of SNPs can be stimulus and context dependent , as has recently been shown for human monocytes , where many regulatory variants display functionality only after pathophysiological relevant immune stimuli [31] . The identified SUPT3H-RUNX2 locus contains two variants , rs12206662 and rs10948155 , which are partially independent of each other . Where rs12206662 is located in the first intron of the RUNX2 gene near the second transcription start site ( the so-called P2 promoter ) , rs10948155 is located more than 500kb away from RUNX2 between CDC5L and SUPT3H . However , rs10948155 is in high linkage disequilibrium with SNPs near in the P2 promoter and SNPs located in chondroblast specific enhancer regions ( S16 and S17 Tables ) . Possibly , these enhancer regions regulate RUNX2 gene expression during endochondral differentiation . RUNX2 ( Runt-related transcription factor 2 ) is a master transcription factor for controlling chondrocyte hypertrophy and osteoblast differentiation [32] . Previous genome-wide association studies have identified variants in the SUPT3H-RUNX2 locus associated with other bone and cartilage related phenotypes including height [14] , bone mineral density [10] and ossification of the posterior longitudinal ligament of the spine [12] . All these previously published loci are independent of the two mJSW SNPs identified in the current study . We hypothesize that the SNPs are located in long-range enhancers , which regulate RUNX2 gene expression during endochondral differentiation via a chromatin-loop mediating protein . We have also identified rs10471753 , with PIK3R1 ( Phosphoinositide-3-Kinase , Regulatory subunit 1 alpha ) as the closest and strongest prioritized gene , related to rs10471753 associated with mJSW . Mutations in this gene are known to cause the SHORT syndrome , which is a rare multisystem disease with several manifestations including short stature , hernias , hyper extensibility and delayed dentition [33] . Taken together with the fact that PIK3R1 is differentially expressed in OA affected cartilage , these results identify PIK3R1 as the most likely causal gene . Another possibility is that not PIK3R1 but rather a long-non-coding RNA ( lncRNA ) , lnc-PIK3R1-4:1 , is causal , since a variant in LD with the lead SNP is located in the predicted transcription start site of this lncRNA potentially affecting its expression . Although conserved in mice and zebrafish , thus far no function has been ascribed to this lncRNA [34] . We confirmed a locus previously associated with cartilage thickness , the DOT1L locus . Our identified SNP , rs11880992 is in high LD with the previously reported SNP rs129827744 , and both are associated with cartilage thickness and hip OA [1] . Despite the previously presented suggestive evidence for involvement of DOT1L in chondrogenic differentiation , DOT1L did not receive a high score in our systematic prioritization study; the gene GADD45B , located in the region 500Kb downstream of the lead SNP , received a similar score . GADD45B is a transcriptional co-factor for C/EBP-β , a master regulator of chondrocyte differentiation [35] . Thus , it remains unclear which gene or genes in this locus contribute to the cartilage phenotype . Further research is needed to determine whether DOT1L is the true causal gene in this locus . Our analyses suggest that the majority of prioritized genes in hip OA associated loci are involved in cartilage and bone developmental pathways; including TGFA , RUNX2 , FGFR3 , PTHLH , COL12A1 and others that seem to affect bone and/or cartilage development such as PIK3R1 and KMT2A We hypothesize that the mJSW and OA associated variants influence gene expression regulation . The dysregulation of these genes and mechanisms during development may , later in life , result in an increased risk for OA . The identified mJSW SNPs are associated with hip OA , but not with knee OA . We have analysed the identified SNPs also for association with knee OA in the TREAT-OA meta-analysis dataset [36] , but found no association . This observation fits in the overall finding that many of the identified genetic loci for OA seem to be site-specific [37] , and support the hypothesis that the aetiology of OA is different in each joint . Nevertheless , this observation can still be a result of low power in the GWAS studies that have been done for OA till now [38] , and final conclusions on this aspect cannot be drawn at this point . This is the first report linking TGFA to human OA most likely by affecting mJSW . It may serve as a new target for future therapies . We have identified multiple mJSW associated loci which have previously been associated with other bone and cartilage related phenotypes such as bone mineral density and height , displaying a possible pleiotropic effect for the analysed traits . It will be important to understand how mJSW and OA associated variants can affect the developmental processes that regulate morphometry of the hip joint , including the formation of articular cartilage . Therefore further expression and functional studies are warranted of genes identified to be associated with hip OA phenotypes .
The participating studies were approved by the medical ethics committees of all participating centres , and all participants gave their written informed consent before entering the study We conducted genome-wide association studies of mJSW for each cohort of the discovery stage: Rotterdam Study I ( RS-I ) , Rotterdam Study II ( RS-II ) , TwinsUK , SOF and MrOS using standardized age- , gender and population stratification ( four principal components ) adjusted residuals from linear regression . Cohort description and details of the single GWAS studies are given in S1 Text and S1 Table . The 6 cohorts used in the discovery stage were combined in a joined meta-analysis using inverse variance weighting with METAL [39] . Genomic control correction was applied to the standard errors and P-values before meta-analysis . SNPs with a P value < = 5×10−6 were selected for replication . The top SNPs for each independent locus were taken for replication in seven studies: the Genetics of Osteoarthritis and Lifestyle ( GOAL ) study , the Chingford study , CHECK ( Cohort Hip & Cohort Knee ) , Genetics osteoARthritis and progression ( GARP ) study , the Genetics of Generalized Osteoarthritis ( GOGO ) , the Johnston County Osteoarthritis Project ( JoCo ) and additionally the Nottingham OA case-control study for association with Hip OA ( see Supplemental material for detailed information of the cohorts ) . Association of the SNPs with mJSW was additionally adjusted for height to test its independence . Secondary analyses included: association of the top SNPs with hip OA using logistic regression analysis ( age and gender adjusted and by study centres an/or relatedness when it was pertinent ) . We used conditional analyses to investigate whether there are any independent signals in the identified associated loci , which were implemented using GCTA-COJO analysis [40] . The mJSW was assessed at pelvic radiographs in anteriorposterior position . The mJSW was measured in mm , along a radius from the center of the femoral head , and defined as the shortest distance found from the femoral head to the acetabulum . Within the Rotterdam Study , we used a 0 . 5 mm graduated magnifying glass laid directly over the radiograph to measure the minimal joint space width of the hip joints [41] . Within SOF and MrOS , a handheld caliper and reticule was used to measured mJSW to the nearest 0 . 1mm between the acetabular rim and proximal head of the femur [42] . For CHECK , mJSW was measured semi-automatic with the Software tool HOLY [43] . Radiographic hip OA was defined in the RS-I , RS-II , RSIII , Twins-UK , Chingford , and JoCo studies using Kellgren and Lawrence ( K/L ) scores . Hip OA cases were defined as a K/L score ≥ 2 on either side of the hip or THR due to OA . Hip OA controls were defined as no THR for OA and K/L score ≤ 1 and JSN ≤ 1 . In MrOS and SOF cohorts , radiographic hip OA case-control was defined by a modified Croft grade , as previously described [44] , where cases were defined as a Croft score ≥ 2 on either side of the hip or THR due to OA and controls were defined as a Croft score ≤ 1 on both sides of the hip and no THR . Hip OA cases in the GOAL and Nottingham OA studies were defined by having THR , and controls were radiographically free of hip OA , as previously described [45] . In GARP , hip osteoarthritis was defined as pain or stiffness in the groin and hip region on most days of the preceding month in addition to femoral or acetabular osteophytes or axial joint space narrowing on radiography or prosthesis due to osteoarthritis . In GOGO , hip OA was defined as KL grade > = 2 , or minimal joint space width < = 2 . 5 mm , or the combination of joint space narrowing grade > = 2 and any osteophyte of grade > = 1 , or history of joint replacement for OA . In JoCo , hip OA cases were defined as KL grade > = 2 or THR in at least one hip . Hip OA controls were defined as KL grade < = 1 in both hips . We have used several available tools and publicly available databases to prioritize genes in known and newly discovered osteoarthritis associated regions . Locus gene sets were constructed by taking a region of 500 Kb upstream and 500Kb downstream of the lead SNP of that locus . We analysed 152 genes in 13 independent loci associated with minimal joint space width in the hip joint ( mJSW ) for 7 loci , hip OA for 4 loci , total joint replacement ( TJR ) for 1 locus and total hip replacement ( THR ) for 1 locus [2] . We analysed the following biological evidence for each gene at all loci; Nearest located genes: Taken from the UCSC genome browser , GRCh37/hg19 [46] . DEPICT gene prioritization: Data-driven Expression-Prioritized Integration for Complex Traits , a novel tool designed to identify the most likely causal gene in a given locus and to gene sets that are enriched in the genetic associations [21] . DEPICT was used to prioritize genes in a 1 MB region around the found SNPs that were significant associated with the osteoarthritis phenotype , taking a region of 500 Kb upstream and 500Kb downstream of the lead SNP of that locus . Gene prioritization analysis was performed to directly investigate functional similarities among genes from different associated regions , significance was defined by false discovery rate ( FDR ≤ 5% ) . GRAIL gene prioritization: Gene Relationships Across Implicated Loci ( GRAIL ) , was used to determine connectively between genes across OA implicated loci based on literature associations [22] . A GRAIL analysis was performed on 10 independent OA associated loci , based on existing literature in PubMed till August 2014 . Mouse gene expression and phenotype: For each investigated gene , expression in mouse bone and/or cartilage tissue during several developmental stages as well as for adult tissue was determined using data from the Jackson lab database ( http://www . informatics . jax . org/ ) . In addition mouse phenotype data was also obtained for each gene . OMIM phenotype: Using the Online Mendelian Inheritance in Man ( OMIM ) database we examined which genes were involved in abnormal skeletal growth syndromes when mutated ( http://omom . org ) . Expression quantitative trait loci: eQTL information was taken from the Blood eQTL browser ( http://genenetwork . nl/bloodeqtlbrowser/ ) and the eQTL browser ( http://www . ncbi . nlm . nih . gov/projects/gap/eqtl/index . cgi ) using the lead SNP in each locus [20] . Non-synonymous variants: Last we determined if there were any nonsynonymous variants in LD ( r2>0 . 06 ) with the lead SNP of a locus , using HaploReg V2 and the SNP Annotation and Proxy Search ( SNAP ) tools [47 , 48] . For each gene we assigned a score based on equally weighted lines of evidence . We have used cartilage samples from the RAAK study to study gene expression in preserved and affected cartilage from individuals undergoing joint replacement [19] . The ongoing Research Arthritis and Articular Cartilage ( RAAK ) study is aimed at the biobanking of blood , joint materials ( cartilage , bone and where available ligaments of knees and hips ) and bone marrow stem cells ( hip joints only ) of patients and controls in the Leiden University Medical Center and collaborating outpatient clinics in the Leiden area . At the moment of collection ( within 2 hours following surgery ) tissue was washed extensively with phosphate buffered saline ( PBS ) to decrease the risk of contamination by blood , and cartilage was collected of the weight-bearing area of the joint . Cartilage was classified macroscopically and collected separately for macroscopically OA affected and preserved regions . Classification was done according to predefined features for OA related damage based on color/whiteness of the cartilage , based on surface integrity as determined by visible fibrillation/crack formation , and based on depth and hardness of the cartilage upon sampling with a scalpel . During collection with a scalpel , care was taken to avoid contamination with bone or synovium . Collected cartilage was snap frozen in liquid nitrogen and stored at -80°C prior to RNA extraction . Tissues have been stored tailored to apply staining and immunohistochemistry ( IHC ) . Furthermore , DNA and RNA have been isolated from the preserved and affected areas of the respective tissues in order to apply genetic , transcriptomic and epigenomic profiling with respect to the OA pathophysiological process . After in vitro transcription , amplification , and labeling with biotin-labeled nucleotides ( Illumina TotalPrep RNA Amplification Kit ) Illumina HumanHT-12 v3 microarrays were hybridized . Sample pairs were randomly dispersed over the microarrays , however each pair was measured on a single chip . Microarrays were read using an Illumina Beadarray 500GX scanner and after basic quality checks using Beadstudio software data were analyzed in R statistical programming language . Intensity values were normalized using the “rsn” option in the Lumi-package and absence of large scale between-chip effects was confirmed using the Globaltest-package in which the individual chip numbers were tested for association to the raw data . After removal of probes that were not optimally measured ( detection P >0 . 05 in more than 50% of the samples ) a paired t-test was performed on all sample pairs while adjusting for chip ( to adjust for possible batch effects ) and using multiple testing correction as implemented in the “BH” ( Benjamini and Hochberg ) option in the Limma-package . Analyses for differential expression between OA and healthy and between preserved and healthy cartilage was performed likewise , adjusting in addition for sex and for age . Exome sequencing was performed in 2628 individuals from the Rotterdam Study the average mean coverage was 55x , corresponding to approximately 80% of the targeted regions covered by at least 20 reads . The exome sequencing was performed in house ( HuGe-F , www . glimDNA . org ) . Details of the technical procedure and variant calling are given in S2 Text . We tested the exome variants for association with mJSW and/or hip OA in two ways . Each individual variant was tested for association with mJSW using the single variant option within RV-test , while adjusting for age and sex . In addition , we did a burden test for each of the selected genes by using SNP-set kernel association test ( SKAT-O ) . SKAT aggregates individual score test statistics of SNPs in a SNP set and computes SNP-set level p-values for a gene [23] . For each of the top mJSW GWAS associated SNPs the LD region was determined using the 1000G Phase 1 population using the Haploreg tool [47] . The LD threshold was set at r2≥0 . 8 . For each of these SNPs it was determined if the variant was located in a potential enhancer region using the Roadmap consortium reference epigenomes data set [27] . Heatmaps were constructed by calculating the percentage of variants in LD with the top mJSW GWAS found SNP located in enhancer regions as defined by the Roadmap epigenome chromatin states . The reference epigenomes were downloaded from the official data portal accompanying [27] . Reference epigenome data was used from mesenchymal stem cell derived chondrocyte cultured cells , Osteoblast , Bone marrow derived cultured mesenchymal stem cells , K562 , HUVEC , HeLA and NHEK cells . Reference epigenomes were chromatin state models based on ChIPseq data of 5 core histone marks ( H3K4me3 , H3K4me1 , H3K36me3 , H3K27me3 , H3K9me3 ) and an additional H3K27ac histone mark , the Roadmap expanded 18-state model . ChIPseq data of mesenchymal stem cell derived chondrocyte cultured cells , and bone marrow derived cultured mesenchymal stem cells were generated by the NHI roadmap epigenomics project [28] . ChIPseq data of , Osteoblast , K562 , HUVEC , HeLA and NHEK cells were generated by the ENCODE consortium [26] . All data and annotation tracks were downloaded through the UCSC genome browser table tool . Visualization of all ChIPseq annotation and roadmap full epigenomes tracks was done through the UCSC genome browser on GRCh37/hg19 . Heatmaps were plotted in R using the CRAN software packages gplots and RcolorBrewer . Enrichment was calculated according to methods described in Trynka et al [25] .
|
Osteoarthritis ( OA ) is the most common form of arthritis and a leading cause of chronic disability in the western society affecting millions of people . OA is a degenerative joint disease characterized by changes in all joint tissues , including cartilage , bone and synovium , causing chronic pain and loss of function . There are no effective therapeutic treatments available for OA and therefore finding novel biological pathways through genetic association studies can open up new treatment options . The number of known DNA variants associated with OA-risk is limited . To identify new loci , we have performed a Genome Wide Association Study meta-analysis on cartilage thickness , one of the joint tissues affected in OA in a total sample of more than 20 , 000 individuals from twelve cohorts . This analysis revealed six variants associated with cartilage thickness , four of these being novel associations , including TGFA as the most prominent one . A systematic prioritization for underlying causal genes , using diverse lines of evidence , highlighted genes underlying the disease associations , including TGFA , RUNX2 and PIK3R1 . Large scale exome sequencing data ( n = 2 , 050 individuals ) indicated that there were no rare exonic variants that could explain the identified associations . This is the first report linking TGFA to human OA , which may serve as a new target for future therapies
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"rheumatology",
"medicine",
"and",
"health",
"sciences",
"pelvis",
"genome",
"analysis",
"cartilage",
"osteoarthritis",
"musculoskeletal",
"system",
"genomics",
"arthritis",
"hip",
"connective",
"tissue",
"biological",
"tissue",
"gene",
"expression",
"joints",
"(anatomy)",
"genetic",
"loci",
"anatomy",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"human",
"genetics"
] |
2016
|
Novel Genetic Variants for Cartilage Thickness and Hip Osteoarthritis
|
Using the guinea pig as a model host , we show that aerosol spread of influenza virus is dependent upon both ambient relative humidity and temperature . Twenty experiments performed at relative humidities from 20% to 80% and 5 °C , 20 °C , or 30 °C indicated that both cold and dry conditions favor transmission . The relationship between transmission via aerosols and relative humidity at 20 °C is similar to that previously reported for the stability of influenza viruses ( except at high relative humidity , 80% ) , implying that the effects of humidity act largely at the level of the virus particle . For infected guinea pigs housed at 5 °C , the duration of peak shedding was approximately 40 h longer than that of animals housed at 20 °C; this increased shedding likely accounts for the enhanced transmission seen at 5 °C . To investigate the mechanism permitting prolonged viral growth , expression levels in the upper respiratory tract of several innate immune mediators were determined . Innate responses proved to be comparable between animals housed at 5 °C and 20 °C , suggesting that cold temperature ( 5 °C ) does not impair the innate immune response in this system . Although the seasonal epidemiology of influenza is well characterized , the underlying reasons for predominant wintertime spread are not clear . We provide direct , experimental evidence to support the role of weather conditions in the dynamics of influenza and thereby address a long-standing question fundamental to the understanding of influenza epidemiology and evolution .
Influenza A virus , of the family Orthomyxoviridae , carries an RNA genome consisting of eight segments of negative-stranded RNA . This genome encodes one or two non-structural proteins and nine structural proteins , which , together with a host cell–derived lipid envelope , comprise the influenza virus particle . Influenza virus causes widespread morbidity and mortality among human populations worldwide: in the United States alone , an average of 41 , 400 deaths and 1 . 68 million hospitalizations [1] are attributed to influenza each year . In temperate regions like the United States , this impact is felt predominantly during the winter months; that is , epidemics recur with a highly predictable seasonal pattern . In northern latitudes , influenza viruses circulate from November to March , while in the southern hemisphere influenza occurs primarily from May to September [2] . Tropical regions , by contrast , experience influenza throughout the year , although increased incidence has been correlated with rainy seasons [2 , 3] . Despite extensive documentation of the seasonal cycles of influenza and curiosity as to their causes , little concrete data is available to indicate why influenza virus infections peak in the wintertime . Theories to explain the seasonal variation of influenza have therefore proliferated over the years ( reviewed in [4] ) . Current hypotheses include fluctuations in host immune competence mediated by seasonal factors such as melatonin [5] and vitamin D [6] levels; seasonal changes in host behavior , such as school attendance , air travel [7] , and indoor crowding during cold or rainy weather; and environmental factors , including temperature [8] , relative humidity ( RH ) , and the direction of air movement in the upper atmosphere [9] . In early studies using mouse-adapted strains of influenza virus , experiments performed in the winter months yielded a transmission rate of 58 . 2%; in contrast , a rate of only 34 . 1% was observed in the summer months [10] . While these data suggested that the seasonal influences acting on humans also affect laboratory mice , no mechanism to explain the observations was identified . Herein , we directly tested the hypotheses that ambient air temperature and RH impact the efficiency with which influenza virus is spread . As a mammalian animal model we used Hartley strain guinea pigs , which we have recently shown to be highly susceptible to infection with human influenza viruses [11] . Importantly , we also found that naïve guinea pigs readily become infected when exposed to inoculated guinea pigs , unlike mice , which do not efficiently transmit influenza virus [11] . Thus , by housing infected and naïve guinea pigs together in an environmental chamber , we were able to assess the efficiency of transmission under conditions of controlled RH and temperature . Our data show that both RH and temperature do indeed affect the frequency of influenza virus transmission among guinea pigs , although via apparently differing mechanisms .
The results of transmission experiments performed at 20 °C and five different RHs ( 20% , 35% , 50% , 65% , and 80% ) indicated that the efficiency of aerosol spread of influenza virus varied with RH . Transmission was highly efficient ( occurred to three or four of four exposed guinea pigs ) at low RH values of 20% or 35% . At an intermediate RH of 50% , however , only one of four naïve animals contracted infection . Three of four exposed guinea pigs were infected at 65% RH , while no transmission was observed at a high RH of 80% ( Figure 2 ) . Where transmission was observed , the kinetics with which infection was detected in each exposed animal varied between and within experiments . To an extent , we believe this variation is due to the stochastic nature of infection . However , while most infection events were the product of primary transmission from an inoculated animal , others could be the result of secondary transmission from a previously infected , exposed guinea pig . With the exception of the lack of transmission at 80% RH , the observed relationship between transmission and RH is similar to that between influenza virus stability in an aerosol and RH [12] , suggesting that at 20 °C the sensitivity of transmission to humidity is due largely to virus stability . To test whether cold temperatures would increase transmission , the ambient temperature in the chamber was lowered to 5 °C and experiments were performed at 35%–80% RH . Overall , transmission was more efficient at 5 °C: 75%–100% transmission occurred at 35% and 50% RH , and 50% transmission was observed at 65% and 80% RH ( Figure 3A–3H ) . The statistical significance of differences in transmission rates at 5 °C compared to 20 °C was assessed using the Fisher's exact test . While at 35% and 65% RH the difference was not found to be significant , at both 50% and 80% RH , transmissibility at 5 °C was found to be greater than that at 20 °C ( p < 0 . 05 ) . Conversely , when the ambient temperature was increased to 30 °C and transmission experiments carried out at a low RH of 35% , no transmission was observed ( Figure 3I and 3J ) . We also observed that , although changes in RH did not affect the kinetics of viral shedding in inoculated guinea pigs , changes in temperature did . At 5 °C and all RHs tested , the intranasally inoculated guinea pigs shed higher titers of virus on days 4 , 6 , and 8 post-infection; most notably , peak shedding was extended in these animals by 2 d relative to guinea pigs housed at 20 °C ( Figure 4 ) . We tested whether the peak duration of viral shedding was statistically longer at 5 °C than at 20 °C by comparing the last day on which the nasal wash titer was ≥ 106 plaque-forming units ( PFU ) /ml for each guinea pig housed at 5 °C and at 20 °C . The average of this last day value was 3 . 93 ± 0 . 63 d for animals at 5 °C and 2 . 21 ± 0 . 61 d for animals at 20 °C . By the Student's t-test for two independent samples , this value was significantly higher for animals housed at 5 °C than those housed at 20 °C ( p < 0 . 0005 ) . The increased duration of peak shedding at 5 °C is most likely the cause of increased transmission under cold conditions . The observed differences in viral titers between animals housed at 20 °C and 5 °C suggested that cold temperature has a negative impact on host defenses early in infection . We therefore used real-time PCR to assess innate immune function in guinea pigs housed at both temperatures . The nasal turbinates were removed from three animals on each of days 1 , 2 , 3 , 5 , and 7 p . i . RNA extracted from these tissues was subjected to reverse transcription with an oligo dT primer followed by quantitative , real-time PCR with target-specific primers . Mx1 , TLR3 , MDA-5 , IRF7 , STAT1 , RANTES , MCP1 , MCP3 , and IL-1β were found to be upregulated in response to infection in animals housed under both conditions ( Figure 5 ) . Peak expression was seen at day 3 p . i . for all of these genes except RANTES , which increased steadily up to day 7 p . i . , and MDA-5 and STAT1 , which showed similarly high levels of expression on days 2 and 3 p . i . While RANTES expression was elevated more in animals at 5 °C than in animals at 20 °C ( p = 0 . 09 ) , peak levels of IL-1β and MDA-5 were higher in guinea pigs at 20 °C ( p < 0 . 05 ) . TLR3 , MCP1 , MCP3 , and MDA-5 were all expressed to significantly higher levels in animals at 5 °C than at 20 °C late in infection ( day 5 and/or 7 p . i . ) , an observation that may simply reflect the levels of virus present at these time points . No significant upregulation was observed in either group of guinea pigs for TNFα , TBK1 , IRF5 , or IFNγ ( Figure 5 ) . Overall , these data indicate that innate immunity was not greatly impaired in guinea pigs housed at 5 °C relative to those at 20 °C . These findings argue against the idea that increased physiological stress experienced under cold conditions leads to a weakening of the immune response [13] .
Three mechanisms could potentially explain the observed influence of RH on transmission . The first acts at the level of the host: breathing dry air could cause desiccation of the nasal mucosa , leading to epithelial damage and/or reduced mucociliary clearance , which would in turn render the host more susceptible to respiratory virus infections . Long-term exposure to dry air is likely to affect influenza virus growth in the upper respiratory tract , and may indeed play a role in influenza seasonality . Nevertheless , based on the brevity of the exposure of naïve guinea pigs to dry air before becoming infected ( less than 72 h ) , we do not believe that this mechanism played a significant role in the observed effects . The second mechanism acts at the level of the virus particle . The stability of influenza virions in an aerosol has been reported to vary with RH [12 , 14 , 15] . The most recent of these reports [12] shows viral stability to be maximal at low RH ( 20%–40% ) , minimal at intermediate RH ( 50% ) , and high at elevated RH ( 60%–80% ) . The similarity between these data on stability and our own on transmission is striking , and suggests that the stability of the virus in aerosols is a key determinant of influenza virus transmission ( with the exception of the absence of transmission at high RH ) . The third mechanism acts at the level of the vehicle , the respiratory droplet . At low RH , evaporation of water from exhaled bioaerosols would occur rapidly , leading to the formation of droplet nuclei; conversely , at high RH , small respiratory droplets would take on water , increase in size and settle more quickly out of the air [16] . Droplet nuclei are less than 5 μm in diameter and , unlike larger droplets , they remain airborne for an extended period of time , thereby increasing the opportunity for transmission of pathogens they carry [17] . Our data suggest that , in this model system , the formation of droplet nuclei is important to transmission; we propose that at high RH ( 80% ) exhaled respiratory droplets settle too rapidly to contribute to influenza virus spread . Based on our data , we present a model in Figure 6 of how transmission is affected by changes in RH . Despite the apparent fitness of animals housed at 5 °C , increased viral shedding under these conditions suggests that improved transmission at low temperature could be due to an effect on the host . This effect may act at the level of primary , physical barriers to infection . Cooling of the nasal mucosa is thought to increase the viscosity of the mucous layer and reduce the frequency of cilia beats [8] . In this way , breathing cold air would slow mucociliary clearance and thereby encourage viral spread within the respiratory tract . An alternative possibility is that , also due to cooling of the mucosal layer , virus residing in the upper airways is more stable when infected animals are kept at 5 °C . Improved persistence of released virus would increase the amount of viable virus shed , and would furthermore augment amplification of virus in the nasal passages through re-infection . The block in transmission at 30 °C and 35% RH could be explained by the opposite effect: warming of the nasal mucosa may lead to more rapid inactivation of virus particles . To our knowledge , we demonstrate for the first time that cold temperatures and low relative humidity are favorable to the spread of influenza virus . Although other factors likely contribute to the periodicity of influenza epidemics , it is clear that air temperature and RH could play an important role . Mathematical modeling indicates that only a small seasonal forcing is required to produce oscillations in infection rate of high amplitude [18]; accordingly , it is possible that the extended exposure of a small proportion of the population to outdoor winter conditions would comprise a sufficient force to create seasonal epidemics . Although the effect may therefore be small and difficult to detect , the importance of RH and temperature in the epidemiology of human influenza could be verified based on surveillance data . To this end , surveillance data with greater spatial resolution ( i . e . , on a regional rather than national scale ) and concurrent meteorological data is needed . The observed lack of transmission among animals housed at 30 °C raises the question of how , if our model is representative of human infection , the spread of influenza viruses occurs in tropical climates . Experiments to more closely examine this question are underway . Our preliminary data suggest that transmission at 30 °C is not improved at RHs higher than 35% . It will be interesting to test whether transmission among guinea pigs housed in the same cage ( i . e . , direct contact or fomite-mediated infection ) is affected by RH and temperature . If transmission in this setting is not impaired at 30 °C , this may suggest that contact-based spread predominates in the tropics , whereas aerosol transmission plays a larger role in temperate climates . Using the guinea pig model , we report a systematic analysis of the effects of RH and temperature on influenza virus transmission in a controlled setting . These data provide valuable insight into the seasonality of influenza and will aid further research into both local and global patterns of influenza virus spread within and between human populations . Our findings furthermore suggest a novel means of infection control for an important human pathogen . Influenza virus transmission indoors could potentially be curtailed by simply maintaining room air at warm temperatures ( >20 °C ) and either intermediate ( 50% ) or high ( 80% ) RHs .
Influenza A/Panama/2007/99 virus ( Pan/99; H3N2 ) was kindly supplied by Adolfo García-Sastre and was propagated in Madin Darby canine kidney cells . Female Hartley strain guinea pigs weighing 300–350 g were obtained from Charles River Laboratories . Animals were allowed free access to food and water and kept on a 12-h light/dark cycle . Guinea pigs were anesthetized for the collection of blood and of nasal wash samples , using a mixture of ketamine ( 30 mg/kg ) and xylazine ( 2 mg/kg ) , administered intramuscularly . All procedures were performed in accordance with the Institutional Animal Care and Used Committee guidelines . During guinea pig transmission experiments , strict measures were followed to prevent aberrant cross-contamination between cages: sentinel animals were handled before inoculated animals , gloves were changed between cages , and work surfaces were sanitized between guinea pigs . The term “aerosol” is used herein to describe respiratory droplets of all sizes . The term “droplet nuclei” is used to refer to droplets that remain airborne ( typically less than 5 μm in diameter ) . Each transmission experiment involved eight guinea pigs . On day 0 , four of the eight guinea pigs were inoculated intranasally with 103 PFU of influenza A/Panama/2007/99 virus ( 150 μl per nostril in phosphate buffered saline [PBS] supplemented with 0 . 3% bovine serum albumin [BSA] ) and housed in a separate room from the remaining animals . At 24 h p . i . , each of the eight guinea pigs was placed in a “transmission cage” , a standard rat cage ( Ancare R20 series ) with an open wire top , which has been modified by replacing one side panel with a wire grid . The transmission cages were then placed into the environmental chamber ( Caron model 6030 ) with two cages per shelf , such that the wire grids opposed each other ( Figure 1 ) . In this arrangement , the guinea pigs cannot come into physical contact with each other . Each infected animal was paired on a shelf with a naïve animal . The guinea pigs were housed in this way for 7 d , after which they were removed from the chamber and separated . On day 2 p . i . ( day 1 post-exposure ) and every second day thereafter up to day 12 p . i . , nasal wash samples were collected from anesthetized guinea pigs by instilling 1 ml of PBS-BSA into the nostrils and collecting the wash in a Petri dish . Titers in nasal wash samples were determined by plaque assay of 10-fold serial dilutions on Madin Darby canine kidney cells . Serum samples were collected from each animal prior to infection and on day 17 post-infection , and seroconversion was assessed by hemagglutination inhibition assay . All transmission experiments reported herein were performed between September 2006 and April 2007 . Guinea pigs were inoculated with 103 PFU of Pan/99 virus intranasally and immediately housed under the appropriate conditions ( 5 °C or 20 °C and 35% RH ) . At days 1 , 2 , 3 , 5 , and 7 post-infection , three guinea pigs were killed and their nasal turbinates removed . Tissues were placed immediately in RNAlater reagent ( Qiagen ) , and stored at 4 °C for 1 to 5 d . RNA was extracted from equivalent masses of tissue using the RNAeasy Protect Mini kit ( Qiagen ) and subjected to DNAse treatment ( Qiagen ) . One microgram of RNA was subjected to reverse transcription using MMLV reverse transcriptase ( Roche ) . One microlitre of the resultant product was used as the template in a SYBR green ( Invitrogen ) real-time PCR assay ( Roche Light Cycler 480 ) with Ampli-taq Gold polymerase ( Perkin-Elmer ) . Primers used were as follows: β-actin f AAACTGGAACGGTGAAGGTG; β-actin r CTTCCTCTGTGGAGGAGTGG; Mx1 f CATCCCYTTGrTCATCCAGT; Mx1 r CATCCCyTTGRTCATCCAGT; MDA-5 f GAGCCAGAGCTGATGARAGC; MDA-5 r TCTTATGWGCATACTCCTCTGG; IL-1β f GAAGAAGAGCCCATCGTCTG; IL-1β r CATGGGTCAGACAACACCAG; RANTES f GCAATGCTAGCAGCTTCTCC; RANTES r TTGCCTTGAAAGATGTGCTG; TLR3 f TAACCACGCACTCTGTTTGC; TLR3 r ACAGTATTGCGGGATCCAAG; TNFα f TTCCGGGCAGATCTACTTTG; TNFα r TGAACCAGGAGAAGGTGAGG; MCP-1 f ATTGCCAAACTGGACCAGAG; MCP-1 r CTACGGTTCTTGGGGTCTTG; MCP-3 f TCATTGCAGTCCTTCTGTGC; MCP-3 r TAGTCTCTGCACCCGAATCC; IFNγ f GACCTGAGCAAGACCCTGAG; IFNγ r TGGCTCAGAATGCAGAGATG; STAT1 f AAGGGGCCATCACATTCAC; STAT1 r GCTTCCTTTGGCCTGGAG; TBK1 f CAAGAAACTyTGCCwCAGAAA; TBK1 r AGGCCACCATCCAykGTTA; IRF5 f CAAACCCCGaGAGAAGAAG; IRF5 r CTGCTGGGACtGCCAGA; IRF7 f TGCAAGGTGTACTGGGAGGT; IRF7 r TCACCAGGATCAGGGTCTTC ( where R = A or G , Y = C or T , W = A or T , K = T or G ) . Primer sequences were based either on guinea pig mRNA sequences available in GenBank ( MCP1 , MCP3 , IL-1b , IFNγ , RANTES , TLR3 , TNFα , and β-actin ) , or on the consensus sequence of all species available in GenBank ( Mx1 , MDA-5 , IRF5 , IRF7 , STAT1 , and TBK1 ) . Sequencing of each PCR product indicated that all primer pairs were specific for the expected transcript . Reactions were performed in duplicate and normalized by dividing the mean value of the cycle threshold ( Ct ) of β-actin expressed as an exponent of 2 ( 2Ct ) by the mean value of 2Ct for the target gene . The fold-induction over the mock-infected was then calculated by dividing the normalized value by the normalized mock value . Data is represented in Figure 5 as the mean of three like samples ( nasal turbinates harvested on the same day p . i . from three guinea pigs ) ± standard deviation . Statistical analyses were performed using GraphPad Prism 5 software .
The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank/index . html ) accession numbers of guinea pig genes used for primer design are as follows: β-actin ( AF508792 . 1 ) ; IFNγ ( AY151287 . 1 ) ; IL-1β ( AF119622 ) ; MCP-1 ( L04985 ) ; MCP-3 ( AB014340 ) ; RANTES ( CPU77037 ) ; TLR3 ( DQ415679 . 1 ) ; and TNFα ( CPU77036 ) .
|
In temperate regions influenza epidemics recur with marked seasonality: in the northern hemisphere the influenza season spans November to March , while in the southern hemisphere epidemics last from May until September . Although seasonality is one of the most familiar features of influenza , it is also one of the least understood . Indoor crowding during cold weather , seasonal fluctuations in host immune responses , and environmental factors , including relative humidity , temperature , and UV radiation have all been suggested to account for this phenomenon , but none of these hypotheses has been tested directly . Using the guinea pig model , we have evaluated the effects of temperature and relative humidity on influenza virus spread . By housing infected and naïve guinea pigs together in an environmental chamber , we carried out transmission experiments under conditions of controlled temperature and humidity . We found that low relative humidities of 20%–35% were most favorable , while transmission was completely blocked at a high relative humidity of 80% . Furthermore , when guinea pigs were kept at 5 °C , transmission occurred with greater frequency than at 20 °C , while at 30 °C , no transmission was detected . Our data implicate low relative humidities produced by indoor heating and cold temperatures as features of winter that favor influenza virus spread .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"viruses",
"in",
"vitro",
"guinea",
"pig",
"virology"
] |
2007
|
Influenza Virus Transmission Is Dependent on Relative Humidity and Temperature
|
Prion diseases are irreversible progressive neurodegenerative diseases , leading to severe incapacity and death . They are characterized in the brain by prion amyloid deposits , vacuolisation , astrocytosis , neuronal degeneration , and by cognitive , behavioural and physical impairments . There is no treatment for these disorders and stem cell therapy therefore represents an interesting new approach . Gains could not only result from the cell transplantation , but also from the stimulation of endogenous neural stem cells ( NSC ) or by the combination of both approaches . However , the development of such strategies requires a detailed knowledge of the pathology , particularly concerning the status of the adult neurogenesis and endogenous NSC during the development of the disease . During the past decade , several studies have consistently shown that NSC reside in the adult mammalian central nervous system ( CNS ) and that adult neurogenesis occurs throughout the adulthood in the subventricular zone of the lateral ventricle or the Dentate Gyrus of the hippocampus . Adult NSC are believed to constitute a reservoir for neuronal replacement during normal cell turnover or after brain injury . However , the activation of this system does not fully compensate the neuronal loss that occurs during neurodegenerative diseases and could even contribute to the disease progression . We investigated here the status of these cells during the development of prion disorders . We were able to show that NSC accumulate and replicate prions . Importantly , this resulted in the alteration of their neuronal fate which then represents a new pathologic event that might underlie the rapid progression of the disease .
Prion diseases or transmissible spongiform encephalopathies ( TSEs ) are fatal neurodegenerative disorders , which include Creutzfeldt-Jakob disease in humans , scrapie in sheep and goats , and bovine spongiform encephalopathy in cattle . Their origin can be genetic , sporadic or infectious and there is currently no available treatment preventing the widespread neurodegeneration occurring in these disorders . TSEs are pathophysiologically characterized by the accumulation in the brain of a pathogenic abnormal isoform of a protein termed PrP scrapie ( PrPSc ) [1] . According to the prion hypothesis , the infectious isoform PrPSc can trigger the autocatalytic conversion of the neuronal host-encoded PrPC into PrPSc [2] through a poorly understood misfolding process [1] , rendering the progression of the disease dependent upon PrP expression . Several studies have reported early , severe and selective loss of GABAergic interneurons in prion diseases [3] , [4] . These specific changes in neuronal subset may underlie some of the clinical symptoms in prions . The diagnosis of these diseases is difficult and often leaves only a short therapeutic window after the appearance of the first clinical signs [5] . Although important efforts have been made to understand the physiopathogenesis of neurodegenerative disorders , Prion diseases are still incurable and new therapeutic approaches such as cell therapy need to be explored . As a matter of fact , the widespread existence of endogenous neural stem cells ( NSC ) in the adult brain [6] , [7] offers hope that these endogenous cells may be harnessed to repair cellular damages caused by brain injuries . During the past decade , several studies have consistently shown that ( i ) NSC reside in the adult mammalian CNS and that ( ii ) adult neurogenesis occurs throughout the adulthood in the subventricular zone ( SVZ ) of the lateral ventricle ( LV ) or the Dentate Gyrus ( DG ) of the hippocampus ( H ) . Accumulating evidences have clearly shown that a large number of newborn neurons can be generated from adult NSC , and integrate into pre-existing neural circuits [8] . Under physiological conditions , adult NSC follow a highly stereotypic differentiation path to generate neurons in the olfactory bulb and the DG . Adult neurogenesis is also highly sensitive to environmental cues , physiological stimuli and neuronal activity , suggesting that the tailored addition of new neurons might serve specific neuronal functions [9] . Endogenous NSC may also provide a cellular reservoir for replacement of cell lost during normal cell turnover but also after brain injury [10] , [11] . In neurodegenerative affections , particularly those involving pathogenic protein misfolding , the field of adult neurogenesis only begins to be explored . The results are not always consistent between studies . For instance , hippocampal neurogenesis is increased in patients with AD [12] , but it is decreased in some transgenic mouse models of AD [13] . Following brain injuries , adult neurogenesis can be increased and is even accompanied by a migration of neural precursors towards the injured area [13] , [14] . However , the activation of this system does not fully compensate the neuronal loss that occurs during diseases and could even contribute to the disease progression [15] . In prion diseases , while it has been suggested that adult neurogenesis was increased [16] , the role and the status of adult NSC are still obscure . Despite the fact that we were able to propagate prions in vitro in NSC from fetal [17] or adult origin , we did not know whether endogenous NSC also accumulated prion in vivo and the impact this would have . This question has been addressed in this study and we showed that endogenous NSC were not only infected by prions but also that their neuronal differentiation process was altered .
In order to investigate whether pathological prion protein ( PrPSc ) deposits were present in brain area containing adult neural stem cells and/or their derivative neuroblasts , we first performed immunohistochemical analyses of PrPSc , nestin and doublecortin in mice that have been intracerebrally infected with the ME7 prion strain . Nestin is a NSC marker and doublecortin ( DCX ) is a neuroblast and immature neurons marker . The PrPSc deposits were densely present in the DG and LV neuronal progenitor area as well as in the DCX neuroblasts area surrounding the LV in mouse infected brain at the endpoint of the disease ( Figure 1A , B , C , E ) . The double immunohistochemical analysis of the nestin or DCX markers with PrPSc clearly confirmed that PrPSc deposits were present around nestin and DCX positive cells ( Figure 1D , F , G , H ) . We then aimed to determine whether endogenous NSC had been infected during the disease development . Adult NSC were therefore isolated from the hippocampus and the lateral ventricle of 10 mock and 10 ME7 prion inoculated mice , at 130 days post-infection ( dpi ) . As expected , the cells cultivated under neurosphere free floating conditions were all , positive for the nestin NSC marker ( Figure 2A ) . We showed after two subpassages ( 30 days after isolation ) that only NSC derived from ME7 infected mice were positive for PrPSc accumulation as assessed by both immunofluorescence ( Figure 2B ) and western blot after Proteinase K digestion ( Figure 2C ) . This PrPSc generation in adult NSC was shown to be stable since PrPSc could still be detected after 15 subpassages ( Figure 2D ) . A thorough control experiment was then designed to confirm that the isolated cells were endogenously infected before the derivation and not during the cell culture . It consisted in the derivation of adult NSC from actin-GFP mice in the presence of an equivalent amount of a 130 dpi ME7 infected brain tissue . To avoid non-actin-GFP cells to proliferate , and therefore obtain only actin-GFP-NSC , the infected tissue dissected in each neurogenesis area was successively frozen at −80°C and heated at +60°C ( Figure 3A ) . This procedure kills all the cells in the extract from non actin-GFP mice . In this paradigm , the derived actin-GFP NSC were not positive for PrPSc ( Figure 3B ) indicating that , in our experimental conditions , the PrPSc particles present in a 130 days infected brain were not sufficient to infect NSC . The differentiation potential of these different NSC isolated from mock and prion infected mice was then analysed . To avoid a massive anoikis which is a ROCK signalling induced apoptosis that occurs after neurosphere dissociation [18] , neurospheres were gently trypsinized and put on polyornithine/laminine coated dishes for one passage before being seeded for differentiation studies . To induce neuronal and glial differentiation , NSC were placed in a differentiation medium during 5 days . In these conditions , the nestin markers completely disappeared ( Figure 4A ) and NSC gave rise to neuroblasts ( Figure 4A ) , young neurons and astrocytes ( Figure 4B ) . Counting analyses of the number of DCX positive cells and betaIII-Tubulin positive cells were performed and are presented in Figure 4 ( Fig . 4A , B , C ) . DCX positive neuroblasts ( 5% and less than 1% for ME7 derived LV-NSC and H-NSC instead of 40% and 25% for non infected derived LV-NSC and H-NSC ) and BetaIII-Tubulin positive new born neurons ( 23 . 5% and 18% for ME7 infected LV-NSC and H-NSC instead of 41% and 40% for non infected LV-NSC and H-NSC ) were less numerous ( p<0 . 05 and p<0 . 01 respectively , MannWhitney Test ) when cells were infected , indicating a defect in neuronal differentiation . Inversely , the proportion of astrocytes was higher in the ME7 infected cells ( Figure 3C ) . In order to assess whether this PrPSc accumulation impaired the neuronal differentiation process itself , NSC cells were infected just when they began their differentiation [17] . Both non infected hippocampus and lateral ventricle derived NSC were plated on poly-L-ornithine/laminin coated dishes in minimal neural N2 medium [17] . They were exposed to ME7 prion homogenate at 0 . 05% ( p/v ) at the beginning of the differentiation process ( simultaneously with the EGF/bFGF withdrawal from the medium ) . N2 medium was completely changed after 24 hours and cells were harvested at different time points to analyse PrPSc accumulation ( Figure 5 ) . PrPSc was able to replicate in both cellular types since PrPSc was detected at 6 dpi for cells derived from the hippocampus ( Figure 5A ) and 8 dpi for NSC derived from the lateral ventricles ( Figure 5B ) . Infected KOPrP cells were used as negative control of the experiment in order to detect the remaining inocula ( Figure 5C ) . These results demonstrated that adult NSC were also capable of replicating ME7 strain during differentiation . Since there was no detectable PrPSc at day 5 of differentiation we analysed the number of neuroblasts ( DCX+ cells , Figure 6A ) and young neurons ( bIII-tubulin+ cells Figure 6B ) after 10 days of differentiation . As observed for the cells that were isolated from prion infected brain , we obtained less DCX+ cells and less βIII-tubulin cells in the infected cells than in the non-infected cells . The difference between non infected and infected cells was also statistically significant for both DCX and βIII-tubulin marker ( P value<0 . 01 , Mann-Whitney Test ) .
The presence of PrPSc among lateral ventricle NSC and neuroblasts incited us to isolate NSC cells from the brain of prion infected mice . We then checked whether these cells were infected by prion or not and assess the impact this could have on their neuronal differentiation . Our results show for the first time that NSC present in prion infected mouse brain accumulate PrPSc , which leads to an alteration of their neuronal fate . Indeed , we isolated adult neural stem cells from neurogenic adult brain area ( the dentate gyrus and the lateral wall of the lateral ventricles ) of both prion infected and non infected mice . Importantly , we were able to keep these cells in culture for several subpassages while maintaining their prion replication . The neuronal differentiation of the infected cells was shown to be compromised since less neuroblasts and less newborn neurons were obtained when the cells were placed in a neuronal differentiation medium . This was also accompanied by an increase in the amount of astrocytes . As neuronal differentiation impairment was also observed when the infection occurred at the beginning of the differentiation , it may suggest that some differentiation pathways are impaired when prion replicate in the cells . This is however not exclusive , with a possible additional impact of PrPSc on cell survival or on cell proliferation , but this remains to be checked in a further study . Indeed , the cellular prion protein has been shown to positively regulate neural precursor proliferation during developmental and adult mammalian neurogenesis [19] , to enhance neural stem cell proliferation [20] and protect cells against oxidative stress [21] , [22] , [23] . A loss/alteration of PrP function could have modified the proliferative potential of the cells resulting in an inappropriate differentiation issue or apoptosis activation as it has been shown with Aβ peptides [15] . Moreover , it has been recently reported [24] that prion neurotoxicity dramatically depends on PrPC expression on established neurons that also appear more susceptible to various subtoxic stimuli such as reactive oxygen species and glutamate . This study suggests also that active prion replication in neurons sensitizes them to environmental stress regulated by neighbouring cells , including astrocytes [24] . Cross-talks between astrocytes and neurons derived from the NSC from infected brains therefore represent an additional mechanism we need to investigate in further studies . In any case , during the development of the brain pathogenesis , the compromised neurogenesis observed may probably have taken place earlier than the onset of hallmark lesions or neuronal loss . We believe that this important alteration of neurogenesis , also suggested to a lesser extent in other proteinopathies [25] , represents an essential mechanism that underlies the progression and the issue of prion diseases . Though adult neurogenesis may be beneficial for regeneration of the nervous system , our results also suggest that this system is defective during the course of the disease . While it has already been shown that neural precursor proliferation was enhanced in prion infected mice [16] the authors did not check the accumulation of PrPSc in the cells of interest . Our study is in fact the first demonstration , in a relevant prion model , of the involvement of neural stem cell in the progression of the disease . Although it also remains to be assessed , adult endogenous neural progenitors could also constitute a “reservoir” of PrPSc amplification since they can proliferate while replicating PrPSc . Moreover , the fact that endogenous adult neurogenesis could be modified by the accumulation of the disease associated misfolded prion protein represents another great challenge . Inhibiting the misfolding of those pathogenic proteins would thus allow the endogenous neurogenesis to compensate injured neuronal system . Our observation regarding the status of neural stem cells during prion infection is also very important since neural stem cells graft approaches are thought to be future therapeutic strategies . This is illustrated in the case of prion diseases by one of our recent publications [26] . In this study the cell therapy approach we developed had a significant effect marked by an increase in both incubation ( 20 . 1% ) and survival times ( 13 . 6% ) in mice grafted before the appearance of the clinical signs . Indeed , the present results suggest that the issue of our preclinical trials would have been more successful if we had proposed a stem cell graft strategy combined with an anti-prion strategy preserving the grafted cells from prion infection and/or targeting the endogenous neural stem cells niches . This is therefore a critical issue in the search for disease-modifying therapies not only for prion disorders but also for other neurodegenerative diseases like Alzheimer or Parkinson diseases .
We have used the NeuroCult Enzymatic Dissociation Kit for Adult Stem Cell ( StemCell Technologies , Grenoble , France ) to isolate adult NSC from ME7 infected and non infected mice . We first dissected the lateral ventricles and hippocampus from adult mouse brains 130 days after their inoculation . They were transferred into a 100 mm dish containing NeuroCult Tissue Collection Solution . The tissue was then chopped with a scalpel for 1 minute and suspended in NeuroCult Dissociation Solution for 7 minutes at 37°C . NeuroCult Inhibition Solution was added at a 1∶1 ratio v/v and the suspension was centrifuged at 700 rpm ( 100× g ) for 7 minutes . Pellet was then resuspended in NeuroCult Resuspension Solution . The digested tissue was mechanically dissociated by pipetting up and down 10 times , and centrifuged at 700 rpm ( 100× g ) for 7 minutes . This step was repeated two more times with a P200 micropipettor . The final pellet was resuspended in 1 mL of Complete NeuroCult NSC Proliferation Medium ( Mouse ) supplemented with 20 ng/mL of recombinant human Epidermal Growth Factor and 10 ng/mL recombinant human basic Fibroblast Growth Factor ( PHG0311 and PHG0021 , Gibco , LifeTechnologies , Saint-Aubin , France ) . The cell suspension was then filtered with a 40 µm cell strainer ( StemCell Technologies , Grenoble , France ) and counted using Trypan Blue . Adult cells were seeded at 3 . 5×103 cells/cm2 in 6-well tissue culture dishes ( Nunc , VWR , Fontenay-sous-Bois , France ) . Under proliferation conditions , adult NSC were cultivated in T-25 cm2 tissue culture flasks ( Nunc , VWR , Fontenay-sous-Bois , France ) . Before differentiation induction , they were first mechanically dissociated and transferred into a Poly-L-Ornithine/laminin coated dish . At 80% of confluence , cells were seeded in Poly-L-Ornithine/laminin wells of 6-well culture dishes with a cell density of 2 . 5×105 cells/cm2 . The day after , the NeuroCult NSC Proliferation Medium was replaced by the NeuroCult NSC Differentiation Medium ( Stem Cell Technologies , Grenoble , France ) . This medium was replaced every 2 days and after 5 days of culture , cells were fixed with 4% paraformaldehyde for 15 minutes at room temperature . Neurospheres were transferred onto Poly-L-Ornithine/laminin coated 6-well plates ( Nunc ) in N2 medium with a cell density of 2 . 5×105 cells/cm2 . They were maintained on monolayer for several subpassages to adapt the cells to the new conditions . When the culture reached the 80–90% of density , 0 . 05% ( p/v ) of ME7 or healthy brain homogenate were added into the N2 medium . Cells were incubated in the presence of the inocula for 24 h . The culture was rinsed twice and fresh N2 medium was added . Media were replaced every 2 days . To monitor any remaining inocula , KOPrP NSC cells were used as control . Cells were permeabilized with 0 . 1% triton X-100 in PBS during 3 minutes , washed with PBS-BSA 0 . 2% three times . Saturation was performed with PBS-0 . 2%BSA for 1 hour at room temperature . Cells were then incubated with the primary antibodies ( Nestin ( Chemicon , MerckMillipore , Billereca , USA ) , DCX ( Abcam , Paris , France ) , GFAP ( Dako , Trappes , France ) , and beta-III tubulin ( TujI clone , Covance , Rueil Malmaison , France ) , 1∶500 in PBS-0 . 2%BSA ) for 1 hour at 37°C . Cells were then washed with PBS-0 . 2%BSA and were incubated with the appropriate secondary antibodies ( goat anti-rabbit-Alexa fluor 488 and goat anti-mouse-Alexa fluor 555 ( Invitrogen LifeTechnologies , Saint-Aubin , France ) , 1∶7000 ) for 1 hour at room temperature . After sequential washes with PBS-0 . 2%BSA , nuclei were stained with Hoechst 33258 ( Calbiochem , VWR , Fontenay-sous-Bois , France ) for 5 minutes under agitation at room temperature and then rinsed with PBS and H2O . The slides were mounted using the FluorSave Reagent mounting medium ( Calbiochem , VWR , Fontenay-sous-Bois , France ) . Photos were taken with a Leica DMRA2 microscope and ImageJ software was used to count the cells . For the statistical analysis we used the Mann-Whitney test . For each condition , images were acquired from 5 to 8 fields and the experiment was repeated three times . An average of 80 cells was counted in each field . Cells were permeabilized with 0 . 5% triton X-100 in PBS during 5 minutes and washed with PBS-BSA 0 . 2% three times . PrPSc epitope retrieval was obtained using 3M guanidium thiocyanate/PBS during 5 minutes and washed with PBS-BSA 0 . 2% and PBS . The saturation was then obtained with PBS-0 . 2%BSA for 1 hour at room temperature . The cells were incubated with the primary antibody SAF61 ( 1∶300 ) in PBS-5% Milk over night at 4°C . The remaining steps were the same as described before [27] . PrPSc presence was checked by western blot analysis using the saf Mix anti-PrP cocktail ( saf 60 , saf 69 and saf 70 antibodies ) as described elsewhere [27] . Mice were anesthetized as described above and then perfused with paraformaldehyde 4% . The brains of the mice were collected and placed in paraformaldehyde 4% for 24 hours at 4°C . They were then manually embedded in paraffin ( Paraplast , Microm , Villefranche sur Saone , France ) and cut in sections of 5 µm thick using a Leica microtome . The sections were collected on microscope slides without treatment ( Starfrost , Microm , Villefranche sur Saone , France ) . Tissues were dewaxed using Clearify solution ( Microm , Villefranche sur Saone , France ) and then rehydrated with decreasing degrees of ethanol washes . Nestin and DCX immunohistochemistry: Sections were incubated in H2O2 0 . 5% for 20 minutes at room temperature and washed with H2O and PBS Epitope retrieval was performed by heathing the sections in 0 . 1 M EDTA . Sections were then saturated with PBS-0 . 1%BSA-0 . 1%Triton X-100 for 1 hour and then incubated overnight at 4°C with the pre-diluted anti-Nestin ( Chemicon , MerckMillipore , Billereca , USA ) primary antibody or anti-DCX primary antibody ( 1/300 , Abcam , Paris , France ) . The secondary antibody used was a biotinylated goat anti-mouse or anti-rabbit antibody ( Amersham , Velizy-Villacoublay , France ) ( 1∶1000 in PBS-0 . 1%triton X-100 ) . The avidin-peroxidase complex ( Vectastain Elite kit , Vector laboratories , Clinisciences , Nanterre , France ) was then added and then revealed with 3 , 3′-diaminobenzidine ( DAB ) ( Vector laboratoriess , Clinisciences , Nanterre , France ) , according to the manufacturers' instructions . PrPSc immunohistochemistry: PrPSc was analysed by immunohistochemistry using the Saf84 ( 0 . 5 µg/ml ) anti-PrP antibody [28] , [29] . SAF84 monoclonal antibody recognising the human 161–170 PrP sequence was kindly provided by Dr J . Grassi ( CEA/SPI , Saclay , France ) . For PrPSc immunostaining , epitope retrieval consists in a treatment with formic acid ( 10 minutes ) followed by an autoclaving treatments ( 121°C , 10 minutes ) . The secondary antibody used was a biotinylated goat anti-mouse antibody ( Amersham , Velizy-Villacoublay , France ) ( 1∶1000 in PBS-0 . 1%triton X-100 ) . The avidin-peroxidase complex ( Vectastain Elite kit , Vector laboratories , Clinisciences , Nanterre , France ) was then added and then revealed with 3 , 3′-diaminobenzidine ( DAB ) . For the double immunostaining procedure , we proceeded as described in [30] . Briefly , we first performed the protocol described above for nestin or DCX immunostaining using EDTA epitope retrieval pretreatments . The slides were then revealed using DAB . The brown precipitate given by DAB resists to PrPSc specific pretreatments ( formic acid and autoclave ) . Then the slides were treated according to the procedure described above for PrPSc immunostaining ( formic acid and autoclave ) . PrPSc revelation was performed using histogreen kits giving a blue green coloration . List of the accession numbers for genes and proteins mentioned in the text ( UniProt ) : PrP: P04925 Nestin: Q6P5H2 Doublecortin: O88809 Beta-III-tubulin: Q9ERD7 Glial fibrillary acidic protein: P03995
|
Prion diseases are irreversible progressive neurodegenerative diseases , leading to severe incapacity and death . They are considered to be caused by an abnormally folded infectious protein named PrPSc . They are characterized in the brain by prion amyloid deposits , vacuolisation , astrocyte proliferation , neuronal degeneration , and by cognitive , behavioural and physical impairments . There is no treatment for these disorders . Transplantation of stem cells , or the stimulation of endogenous neural stem cells ( NSC ) in the adult brain therefore constitute new interesting and promising strategies . While our first interest was the development of a cell therapy approach , we rapidly realised that there were a lot of questions to address before investigating pre-clinical cell therapy assays . Some of them were: ( i ) what is the status of endogenous neural stem cells during the development of prion diseases and can they amplify prions , and ( ii ) can they still proliferate and give rise to new neurons . In this study we definitely demonstrate that PrPSc was not only able to replicate in adult neural stem cells derived from infected brains but also that this results in an impairment of the production of their neuronal derivatives .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"mental",
"health",
"biology",
"neuroscience"
] |
2013
|
Prion Replication Occurs in Endogenous Adult Neural Stem Cells and Alters Their Neuronal Fate: Involvement of Endogenous Neural Stem Cells in Prion Diseases
|
Intragenomic conflicts arise when a genetic element favours its own transmission to the detriment of others . Conflicts over sex chromosome transmission are expected to have influenced genome structure , gene regulation , and speciation . In the mouse , the existence of an intragenomic conflict between X- and Y-linked multicopy genes has long been suggested but never demonstrated . The Y-encoded multicopy gene Sly has been shown to have a predominant role in the epigenetic repression of post meiotic sex chromatin ( PMSC ) and , as such , represses X and Y genes , among which are its X-linked homologs Slx and Slxl1 . Here , we produced mice that are deficient for both Sly and Slx/Slxl1 and observed that Slx/Slxl1 has an opposite role to that of Sly , in that it stimulates XY gene expression in spermatids . Slx/Slxl1 deficiency rescues the sperm differentiation defects and near sterility caused by Sly deficiency and vice versa . Slx/Slxl1 deficiency also causes a sex ratio distortion towards the production of male offspring that is corrected by Sly deficiency . All in all , our data show that Slx/Slxl1 and Sly have antagonistic effects during sperm differentiation and are involved in a postmeiotic intragenomic conflict that causes segregation distortion and male sterility . This is undoubtedly what drove the massive gene amplification on the mouse X and Y chromosomes . It may also be at the basis of cases of F1 male hybrid sterility where the balance between Slx/Slxl1 and Sly copy number , and therefore expression , is disrupted . To the best of our knowledge , our work is the first demonstration of a competition occurring between X and Y related genes in mammals . It also provides a biological basis for the concept that intragenomic conflict is an important evolutionary force which impacts on gene expression , genome structure , and speciation .
Transmission distorters ( TDs ) , also known as segregation distorters or meiotic drivers , are genetic elements that are transmitted to the next generation with a higher frequency than the expected 1∶1 Mendelian inheritance ratio . TDs have the tendency to accumulate in low recombination regions where tight linkage allows cooperation between TDs and responder genes to evolve , as seen in the mouse t-complex [1] [for recent reviews see [2] , [3]] . The non-recombining region of the heteromorphic sex chromosomes is the largest genomic example of recombination suppression [4] , with the consequent potential for TDs to arise and distort the population sex ratio . Theory predicts that an unlinked suppressor of the sex ratio distortion ( whether autosomal or on the other sex chromosome ) would rapidly be selected for to restore the Fisherian 1∶1 sex ratio [5] . A subsequent evolutionary arms race between the distorter and its suppressor may follow and lead to repeated bouts of amplification of the genes involved in this intragenomic conflict [6] . In Drosophila , the X- and Y-encoded multicopy genes Stellate and Suppressor of Stellate are believed to illustrate the genomic conflict theory since deletions of Su ( Ste ) locus lead to a derepression of Stellate associated with a distorted sex ratio towards an excess of females; but to date it remains unclear whether or not Stellate is a transmission distorter [7] , [8] . Intragenomic conflicts over sex chromosome transmission are predicted to have influenced genome structure , gene expression and speciation [2] , [3] . Several cases of sex chromosome transmission distortion have been reported in the literature but they mostly concern Drosophila species [2] , [9]–[14] and remain poorly characterized in mammals . Sex ratio segregation distortion may be more frequent than observed as the distortion is often masked by the presence of a suppressor in wild-type ( WT ) populations [2] , [9]–[15] . In the mouse , the existence of an intragenomic conflict between X- and Y-linked genes has long been suggested: males with a partial deletion of the male specific region of the Y long arm ( MSYq ) produce offspring with a sex ratio skewed towards females [16] , suggesting that MSYq encodes a factor ( s ) suppressing sex ratio distortion . MSYq consists of multicopy gene families , present in ∼60 to 100 copies [17]–[24] , many of which possess X-linked multicopy homologous genes [20] , [25] , [26] . This has been considered a manifestation of a conflict between an X-encoded TD and a Y-encoded suppressor that remain to be identified [16] , [20] , [26] , [27] . We have previously shown that the MSYq-encoded multicopy gene Sly ( Sycp3-like Y-linked ) represses the postmeiotic expression of X and Y genes [19] . Sly-deficient males – also known as shSLY males since they carry a short hairpin RNA-expressing transgene which triggers the specific degradation of Sly transcripts by RNA interference – present a remarkable up-regulation of sex chromosome genes in postmeiotic germ cells ( spermatids ) associated with a loss of repressive epigenetic marks , such as trimethylated histone H3 ( H3K9me3 ) and CBX1 [19] . SLY therefore limits sex chromosome expression via the recruitment/maintenance of repressive epigenetic marks to post meiotic sex chromatin ( PMSC ) and has been proposed to associate with the sex chromosomes through its Cor1 domain – a domain thought to mediate chromatin interactions ( Conserved Domain Database from the National Center for Biotechnology Information , http://www . ncbi . nlm . nih . gov/Structure/cdd/cddsrv . cgi ? uid=147120 ) . Interestingly , Slx and Slxl1 , two multicopy X-linked genes related to Sly [25] have been co-amplified with Sly during the evolution of the mouse genome [18] , [20] and are among the genes that are up-regulated when Sly expression is reduced/absent [19] . Using a strategy of transgenically-delivered short hairpin RNA similar to the one previously used to disrupt the function of Sly , we have recently produced Slx/Slxl1-deficient mice ( also known as shSLX mice ) . This study has shown that Slx/Slxl1 are indispensable for normal sperm differentiation , and that Slx/Slxl1 deficiency leads to the deregulation of a number of autosomal genes [28] . Moreover , both SLY and SLXL1 proteins have now been shown to interact with the acrosomal protein DKKL1 [21] , [29] . In the present study we show that SLX/SLXL1 and SLY proteins have antagonistic effects on gene expression for both the sex chromosomal genes deregulated in shSLY and the set of autosomal genes deregulated in shSLX , and furthermore have antagonistic effects on offspring sex ratio . Our data demonstrate that Slx/Slxl1 and Sly are involved in a postmeiotic intragenomic conflict; we propose this phenomenon has had a strong impact on the structure and epigenetic regulation of the sex chromosomes , and may also have influenced the evolution of hybrid sterility in the mouse lineage .
In normal males , SLX/SLXL1 proteins are located in the cytoplasmic compartment of spermatids [28] , whereas SLY is additionally detected in the spermatid nucleus where it has been shown to colocalize with the X and Y chromosomes [19] . When performing immunofluorescence detection of SLX/SLXL1 proteins on spermatids devoid of SLY protein ( i . e . on shSLY testicular sections ) , we observed an augmented SLX/SLXL1 signal in the cytoplasm compared to controls ( WT ) – confirming up-regulation at the protein level – and some signal in shSLY round spermatid nuclei that was not visible in WT ( Figure 1A ) . The presence of SLX/SLXL1 proteins in shSLY spermatid nuclei was confirmed by Western blot analyses of nuclear fractions ( Figure 1B ) . We then investigated in more detail the nuclear localization of SLX/SLXL1 in the context of Sly deficiency . The vast majority of shSLY spermatid nuclei showed a strong SLX/SLXL1 signal ( 280/369 , 76% ) ( Figure 1C and Figure S1 ) . This signal colocalized with the postmeiotic sex chromatin ( PMSC , i . e . the X or the Y chromosome since spermatids are haploid ) in 96 . 5% of round spermatids ( 82/85; 32/34 for X-bearing and 50/51 for Y-bearing spermatids ) . In comparison , 84% of WT round spermatid nuclei ( 265/316 ) did not have any SLX/SLXL1 signal . The nuclear SLX/SLXL1 signal observed in the remaining ∼16% of WT round spermatid was very weak when compared to the nuclear signal in shSLY round spermatids but appeared to colocalize with the PMSC in the majority of the cases ( Figure S1 ) . In addition to colocalizing with the PMSC , foci of SLX/SLXL1 proteins were observed outside the sex chromatin , reminiscent of the SLY signal present in the nucleus of WT spermatids [19] . We have since established that these ‘ectopic’ SLY sites include a ∼14 Mb cluster of 7 Speer genes on chromosome 5 that are up-regulated in shSLY spermatids . As a result , SLY immunofluorescence followed by fluorescent hybridization of a Speer DNA probe ( DNA FISH ) showed that , in the majority of WT round spermatids ( 107/136 , 78 . 5% ) , SLY protein colocalized with the Speer DNA FISH signal ( Figure 1D ) . We next looked at SLX/SLXL1 proteins in Sly-deficient round spermatids and observed that they colocalized with the Speer gene cluster in 73% of the cases ( 130/178 ) ( Figure 1D ) . Thus , in the absence of SLY , SLX/SLXL1 proteins colocalize with the sex chromatin and with the autosomal Speer gene cluster , mimicking the pattern observed for SLY protein in WT spermatids . We then wondered if the localization of SLX/SLXL1 proteins to the PMSC in the absence of SLY also affects postmeiotic sex chromosome gene expression . To address this question , we generated males that were deficient for SLX/SLXL1 and SLY proteins: we produced males carrying shSLY ( Sly specific short hairpin RNA ) transgene [19] together with one or two shSLX ( Slx/Slxl1 specific short hairpin RNA ) transgenes , shSLX1 and/or shSLX2 [28] . Firstly , we checked the efficiency of Slx/Slxl1 and Sly knockdowns in round spermatids from males carrying shSLX1 and shSLY transgenes ( hereafter named shSLX1shSLY males ) . The reduction in Slx/Slxl1 transcript level was similar in shSLX1shSLY males and in shSLX1 siblings , while Sly knockdown was even stronger in shSLX1shSLY males compared to shSLY siblings ( Figure 2A and 2B ) . Sly transcript quantification included both alternative splice variants ( Sly1 and Sly2 ) [21] which were knocked-down with the same efficiency [19] . No SLY1 protein could be detected in shSLY or in shSLX1shSLY tissues ( Figure 2D–2E ) . To date it remains unclear whether Sly2 transcripts are translated since anti-SLY1 antibody cannot detect SLY2 protein [21] . The discrepancy between transcript and protein levels is likely due to the presence of non-coding Sly transcripts , as previously observed [19] . Reduction in SLX and SLXL1 proteins was similar in shSLX1shSLY males and in shSLX1 siblings ( Figure 2C ) . We also produced shSLX1/2shSLY males that carry the two shSLX transgenes along with the shSLY transgene . As expected , shSLX1/2shSLY males showed a very efficient knockdown of Slx and Slxl1 ( Figure S2A ) ; Sly knockdown in these males was similar to that in shSLX1shSLY males ( Figure S2B ) . Thus , the combination of shSLX and shSLY transgenes gives an efficient knockdown of Slx/Slxl1 and Sly genes; the resulting transgenic males are therefore deficient for Slx/Slxl1 and Sly transcripts and proteins ( hereafter named Slx/y-deficient males ) . We then performed microarray transcriptome analyses on Slx/y-deficient purified round spermatids and compared these results to those obtained from Sly-deficient and from WT round spermatids ( Figure 3 and Figure S3 ) . The up-regulation of X- and Y-encoded spermatid transcripts was significantly less pronounced in Slx/y-deficient males than in Sly-deficient males ( Figure 3A–3C ) . Specifically , 222 genes showed a greater than 1 . 5 fold-increase in Sly-deficient spermatids relative to WT , and 196 of them were corrected to some degree by the additional Slx/Slxl1 deficiency ( i . e . in Slx/y-deficient spermatids ) . As a Y-encoded gene , Sly itself is affected by Slx/Slxl1 knockdown and thus is expressed at a lower level in Slx/y-deficient males than in Sly-deficient males ( Figure 2B and Figure S3 ) . The microarray findings were confirmed for several representative X and Y genes by real time PCR ( Figure 3B and Figure S2C ) . These opposite effects of Sly and Slx/Slxl1 deficiency show that , in the absence of SLY protein , SLX/SLXL1 proteins localize to PMSC where they increase sex chromosome gene expression; when both SLX/SLXL1 and SLY proteins are reduced/absent in PMSC ( in Slx/y-deficient males ) , the level of X- and Y- encoded transcripts is closer to the WT value . It is worth noting that while Slx/Slxl1 deficiency significantly reduces the up-regulation of XY genes induced by Sly deficiency , it does not bring expression all the way back down to WT levels . This may indicate that Slx/Slxl1 knockdown is not sufficient to fully compensate for the effect of Sly deficiency; alternatively it may be that in the WT MF1 laboratory strain , the combined effect of the presence of both SLX/SLXL1 and SLY is a net reduction of XY expression level , thus leading to a net increase when both genes are deficient . The up-regulation of X- and Y-encoded spermatid genes in Sly-deficient spermatids has been shown to be concurrent with a diminution of the repressive epigenetic marks ( such as H3K9me3 ) normally associated with PMSC [19] . We therefore decided to study these repressive marks in Slx/y-deficient spermatids , and observed that H3K9me3 staining on PMSC ( as compared to H3K9me3 chromocenter staining ) was significantly higher ( p = 0 . 00003 ) in Slx/y-deficient spermatids than in Sly-deficient spermatids ( average staining intensity: 0 . 59 and 0 . 51 respectively ) , and closer to but significantly different from the WT value ( average staining intensity in WT: 0 . 65 , p = 0 . 003 ) ( Figure 3D and Figure S4 for quantification ) . Therefore Slx/Slxl1 deficiency partially compensates the loss of H3K9me3 marks induced by Sly deficiency . These results correlate with the global effect of Slx/Slxl1 transcript knockdown on sex chromosome expression and suggest that SLX/SLXL1 and SLY proteins compete in spermatids for access to PMSC where they have activator and repressive effects respectively , at the whole-chromosome level . We then compared the transcriptomes of WT , Slx/Slxl1-deficient and Slx/y-deficient spermatids . This revealed a 10% reduction in Y transcription in Slx/Slxl1-deficient spermatids compared to WT that was not seen in an earlier study [28] ( Figure 3A–3B ) . This reduction is congruent with our observation of some SLX/SLXL1 proteins in a small number of WT spermatid nuclei ( Figure 1B and Figure S1 ) ; this small fraction of SLX/SLXL1 proteins most likely increases sex chromosome gene expression in the nucleus of WT spermatids , while the loss of these proteins leads to a slight reduction of XY expression in Slx/Slxl1-deficient spermatids . A faint reduction of expression was observed for some X genes ( for instance Actrt1 , see Figure 3B ) but this did not significantly differ from the WT value . We have previously shown that Slx/Slxl1 deficiency leads to delay in spermatid elongation and sperm release , associated with the deregulation ( principally the up-regulation ) of 115 genes , the majority of which are located on the autosomes . Given that SLX/SLXL1 proteins are almost entirely cytoplasmic in wild type , we proposed that these transcriptional changes were a manifestation of “cytoplasmic” defects , rather than a direct effect of SLX/SLXL1 proteins on autosomal gene expression; for instance , an as yet unidentified cytoplasmic partner of SLX/SLXL1 could mediate the transcriptional changes that are necessary for normal spermatid elongation , or it may be that the transcriptional changes seen reflect an altered cellular proportion of different step spermatids in shSLX [28] . In the present study , we compared microarray results from Slx/Slxl1-deficient and Slx/y-deficient spermatids and , surprisingly , observed that most of the genes deregulated by Slx/Slxl1 deficiency were less affected in Slx/y-deficient spermatids ( 111/115 genes , Figure 3E and Figure S5 ) . Therefore , Sly knockdown corrects the deregulation of autosomal genes induced by Slx/Slxl1 ( with autosomal gene expression values close to WT levels in Slx/y-deficient spermatids; Figure 3E and Figure S5 ) . These results show that SLX/Y proteins have opposite regulatory effects on autosomal gene expression as well as on sex chromosome gene expression . Our microarray results demonstrate that the deregulation of sex chromosome-linked or autosomal genes observed in Sly-deficient or in Slx/Slxl1-deficient spermatids respectively , is corrected in Slx/y-deficient spermatids; we therefore compared the reproductive parameters of Slx/y-deficient males with those from males that are singly deficient for either Slx/Slxl1 or for Sly . Firstly , Slx/y-deficient males had significantly improved sperm numbers ( Table 1 ) . This was particularly striking for the comparison between shSLX1/2 and shSLX1/2shSLY males: shSLX1/2 males had dramatically reduced spermatozoa numbers but the addition of shSLY transgene to this genotype increased the number of sperm produced ∼50-fold ( Table 1 ) . Low sperm count in shSLX males was attributed to the apoptosis of delayed elongating spermatids [28] . We therefore analyzed spermatid elongation delay and apoptosis in Slx/y-deficient males and in their Slx/Slxl1-deficient siblings . Remarkably , while Slx/Slxl1-deficient males presented a high number of delayed and apoptotic elongating spermatids , Slx/y-deficient models did not significantly differ from WT ( Figure 4A–4B ) . The spermatozoa morphology of Slx/y-deficient males was also much improved compared to that of Slx/Slxl1- or Sly-deficient males ( Figure 4C and Figure S6 ) . Finally , we compared the fertility of Slx/y-deficient males with Slx/Slxl1- or Sly-deficient siblings: Slx/y-deficient males had overall better fertility than males that are deficient for either Slx/Slxl1 or Sly , with reproductive parameters close to WT values ( Table 1 ) . Strikingly , the addition of Sly deficiency was able to reverse the sterility observed in Slx/Slxl1 deficient-males ( line shSLX1/2 ) ( Table 1 ) . All in all , males that were deficient for both Slx/Slxl1 and Sly had considerably better reproductive parameters than males that were deficient for Slx/Slxl1 or Sly alone . These analyses show that Sly deficiency almost completely rescues the defects and gene deregulation induced by Slx/Slxl1 deficiency , while Slx/Slxl1 knockdown only partially rescues those subsequent to Sly deficiency . This may be due to a different knockdown efficiency: indeed , no SLY1 protein can be detected in Slx/y-deficient samples while some SLX/SLXL1 proteins remain ( Figure 2C–2E ) . We previously reported a tendency of an excess of females in the progeny of Sly-deficient males ( 7 . 7% excess of females , Chi-square p = 0 . 0569 ) [19] . While analyzing the fertility of our transgenic lines , we observed that Slx/Slxl1-deficient males ( i . e . shSLX1 ) yielded an offspring sex ratio of 40% ( 74/187 ) female progeny , compared to a ratio of 51% ( 234/461 ) in WT siblings . This represents a statistically significant sex ratio distortion of 11% in favour of male offspring ( Chi-square p = 0 . 006 ) . Importantly , a normal sex ratio was restored by the addition of Sly deficiency: Slx/y-deficient males produced an offspring sex ratio of 50% ( 103/208 ) female progeny that did not differ from WT and was also significantly different from the offspring sex ratio of shSLX1 males ( Chi-square p = 0 . 03 ) . These data show that both Slx/Slxl1 and Sly affect the transmission of X- and Y-bearing gametes , Slx/Slxl1 favouring X transmission while Sly favours Y transmission .
In recent years , we have identified the transcriptional consequences of Sly and Slx/Slxl1 deficiency , and related these to the observed phenotypes in terms of spermatid development , sperm morphology and offspring sex ratio [19] , [28] . Remarkably , we now show that in dual shRNA knockdown models where both genes are deficient , the transcriptional and phenotypic consequences of the individual knockdown are dramatically ameliorated , correcting the X/Y/Speer up-regulation and sperm shape abnormalities seen in Sly-deficient spermatids; the autosomal gene up-regulation , spermatid elongation delay and apoptosis , and sperm shape abnormalities seen in Slx/Slxl1-deficient spermatids; and improving fertility in both cases . Strikingly , however , two different and almost entirely non-overlapping sets of genes are affected by the mutual antagonism of SLX/SLXL1 and SLY . In this discussion , we refer to “Group 1” genes as the set of X/Y/Speer genes up-regulated in Sly-deficient spermatids and ( partially ) corrected in the dual knockdown , and “Group 2” genes as the set of metabolism-related autosomal genes up-regulated in Slx/Slxl1-deficient spermatids and ( almost fully ) corrected in the dual knockdown . Sly regulates the epigenetic repression of post meiotic sex chromatin ( PMSC ) and a few specific autosomal genes such as the Speer cluster . In the nucleus , SLY appears to act via the recruitment/maintenance of the repressive heterochromatin marks CBX1 and H3K9me3 , which consequently limits the expression of X and Y genes in spermatids , among which are its X-linked homologs Slx and Slxl1 [19] . Here , we show that , in the absence of SLY , SLX/SLXL1 proteins relocate to the nuclear sites ( both sex-linked and autosomal ) vacated by SLY proteins . It is unlikely that SLX/SLXL1 nuclear localization in Sly-deficient spermatids is solely a consequence of increased SLX/SLXL1 protein abundance , since there is no clear enrichment in nuclear SLX/SLXL1 proteins in spermatids of transgenic mice overexpressing SLX or SLXL1 ( our unpublished preliminary data ) . Moreover , in the double transgenic model ( Slx/y-deficient males ) where SLX/SLXL1 family members are also reduced/absent , XY gene expression , Speer expression and the intensity of H3K9me3 marks on the sex chromatin are closer to normal values . This indicates that SLX and/or SLXL1 have consequences both for transcriptional activity and for histone modification when present on sex chromatin , and that these are directly opposed to the effects of SLY . We therefore propose that SLX/SLXL1 and SLY proteins compete for access to nuclear sites in spermatids , where they act as positive and negative transcriptional regulators respectively . We cannot at this point say precisely where the competition occurs: it may be directly at the level of chromatin binding within the nucleus , or SLX/SLXL1 and SLY may compete for access to factors affecting nuclear import . We note that SLX and SLXL1 proteins lack nuclear localization signals ( NLS ) while SLY NLS is mutated/truncated [22] , [25]; as such they probably depend on other interacting factors to enter the nucleus . It also remains possible that the competition is mediated indirectly: for example , SLY could affect SLX/SLXL1 intracellular localization via regulating the expression of a third factor controlling SLX/SLXL1 access to the nuclear sites . Slx/Slxl1 deficiency has been shown to increase the level of ∼100 autosomal transcripts which code for proteins of the cytoskeleton and the extracellular matrix , or are implicated in various cytoplasmic processes ( i . e . energy production , lipid metabolism , ubiquitin-mediated degradation , etc . ) [28] . These transcriptional effects are corrected in Slx/y-deficient males , suggesting that these changes may also be manifestations of the same nuclear/chromatin regulatory antagonism exhibited by Group 1 genes , perhaps via relocation of repressive factors from sex chromatin to autosomal locations and vice versa . There are , however , three significant objections to this interpretation . Firstly , as noted previously , in WT spermatids SLX/SLXL1 are predominantly cytoplasmic proteins , and the levels in the nucleus are almost undetectable: it is hard therefore to see i ) how Slx/Slxl1 knockdown could directly induce widespread transcriptional changes , ii ) what would then be the function of the abundant SLX/SLXL1 proteins in the cytoplasm . Secondly , this interpretation would require not only that SLX/SLXL1 act simultaneously as transcriptional activators of Group 1 genes and as transcriptional repressors of Group 2 genes , but that SLY has the reverse effect in both cases: it is challenging to imagine a mechanism that could explain this . Thirdly , if both Group 1 and Group 2 gene effects are a manifestation of the changing balance of SLX/Y proteins in the nucleus and/or of a relocation of repressive factors from the sex chromosomes to autosomes , then both groups of genes would be expected to change together . This is not the case: Group 1 genes are affected in shSLY but not in shSLX , and Group 2 genes vice versa . For this reason , we favour our existing interpretation that Group 2 gene deregulation is a manifestation of the spermiogenesis defects occasioned by cytoplasmic Slx/Slxl1 deficiency ( i . e . spermatid elongation delay and apoptosis , reduced sperm count , abnormal head to tail connections of the spermatozoa and male infertility ) [28] , and is not a direct effect of SLX/SLXL1 proteins on autosomal gene transcription . Given that the ( cytoplasmic ) spermiogenesis defects are corrected in the dual mutant , it stands to reason that the secondary expression changes follow the same pattern . We therefore propose that , in addition to the nuclear effects on Group 1 genes , SLY protein has a cytoplasmic role , opposing that of SLX/SLXL1 . SLY proteins have been shown to be present in both the spermatid nucleus and cytoplasm [19] , [21] . Intriguingly , a recent report indicates that the acrosomal ( cytoplasmic ) protein DKKL1 , which we previously identified as a binding partner of SLY1 [21] , also interacts with SLXL1 [29] . We have performed additional experiments and observed that all SLX/Y family members ( i . e . SLY1 , SLY2 , SLX and SLXL1 ) can interact with DKKL1 ( Figure S7 ) . Therefore , SLX/SLXL1 and SLY proteins could compete for interaction with ( a ) common partner ( s ) in the cytoplasm , and this competition could be at the basis of the opposite effects of SLX/SLXL1 and SLY on spermiogenesis and autosomal gene expression . A combined model proposing how SLX/SLXL1 and SLY proteins have antagonistic effects in both the spermatid nucleus and cytoplasm is presented in Figure 5 . We recognize that under our preferred model , it is difficult to explain the directionality of the expression changes seen in shSLX relative to WT , which was predominantly up-regulation of autosomal genes with comparatively few down-regulated genes [28] . A potential explanation for this lies in the spermatid developmental delay resulting in delayed spermatid elongation in shSLX . This could potentially skew the round spermatid population in shSLX testes towards earlier stages , i . e . proportionally more step 1 spermatids and fewer step 7–8 spermatids . Since there is a progressive transcriptional shutdown throughout spermatid development as chromatin is repackaged in preparation for nuclear condensation , this would thus manifest in shSLX as a selective up-regulation of those genes expressed specifically in early stage round spermatids ( which in turn is plausible given the annotated functional categories for these Group 2 genes ) . Testing this interpretation will require further experiments on fractionated , staged sub-populations of round spermatids . Irrespective of the precise molecular mechanism ( s ) underlying the antagonistic effects of SLX/SLXL1 and SLY , our results demonstrate that both genes have an effect on offspring sex ratio . In particular , comparing shSLX ( where Sly is still present ) to the dual knockdown , there is a significant excess of males; and when comparing shSLY ( where Slx/Slxl1 are still present ) to the dual mutant , there is a trend towards excess of females . Thus , the net effect of these genes on inheritance is for X-linked family members to favour X chromosome transmission , and Y-linked members to favour Y chromosome transmission , constituting a prima facie genomic conflict . Such a conflict was first postulated in the 1990s following observations that male mice with a partial deletion of the Y long arm produce an excess of female offspring , however supporting evidence has not been forthcoming until recently [16] , [20] , [26] , [27] . The present study demonstrates that such a conflict exists between the sex chromosome-linked Sycp3-related genes . An intragenomic conflict is often not visible under normal conditions ( i . e . in a WT population ) [11] , [13] and here the positive effect of Slx/Slxl1 on sex chromosome transcription was uncovered by the production of mice that are deficient for both Sly and Slx/Slxl1; similarly , the effects of Slx/Slxl1 deficiency are also corrected in the dual mutant , although the molecular mechanisms involved are less clear . Under the distorter/responder model exemplified by the t complex [1] , both Slx/Slxl1 and Sly are transmission distorters in that changes in their expression levels lead to a distortion of the sex ratio . However , it is unlikely that they are directly responsible for mediating the transmission skew ( i . e . responder genes ) . Indeed , the physiological mechanism of the skew in the present model is an asymmetry in fertilizing ability between X and Y sperm [30] . This implies an underlying molecular/functional asymmetry , namely the presence of a responder gene product which is not evenly shared between X and Y sperm . Both of the known mammalian examples of transmission ratio distortion depend on non-sharing of gene products ( both transcript and protein ) between sister spermatids: Spam1 in the case of Rb ( 6 . 16 ) and Rb ( 6 . 15 ) translocation heterozygotes , and TcrSmok in the case of driving t haplotypes [31]–[33] . We note that SLX/Y proteins appear to be similarly expressed in X- and Y-bearing spermatids . It therefore seems likely that the distortion in Yq deleted mice and in shSLXshSLY transgenic models is mediated by an as yet unidentified sex-linked gene or gene ( s ) ( i . e . the responder ) , for which Slx/Slxl1 and Sly are competing regulators via their global effects on sex chromatin expression . Among the deregulated genes , a few appear as promising candidates , such as the X-encoded homolog of Tcp11 , which is one of the genes involved in the t-complex transmission distortion , albeit as a distorter rather than a responder [34] , and Alkbh7 , since another Alkbh gene has recently been found to cause sex ratio distortion [35] . However , there may be several linked genes involved , at least one of which is likely to evade transcript sharing . In view of this possibility , it is worth noting that both regulators of the conflict have a global effect on sex chromatin; this is an efficient way to control multiple sex chromosome-linked distorters and/or responders simultaneously . The ease of identifying the responder ( s ) will depend on how directly SLX/Y regulate them and how many there are . Finally , it is possible that autosomal factors also contribute to the regulation of sex-linked transmission distortion . We note that historically , Slx appeared on the X before Sly appeared on the Y , and its distorting effect on sex ratio may have subsequently been countered by a combination of Sly-mediated repression and other autosomal genes being selected to favour a balanced sex ratio [20] . In the mouse lineage , there has been a remarkable amplification of spermatid-expressed sex chromosome genes ( all of which fall into Group 1 identified above ) , and which has had a dramatic influence on the structure of the mouse sex chromosomes . This expansion occurred subsequent to the appearance of Sly , but was not accompanied by a matching increase in XY transcript levels [20] . It is therefore very likely that essential sex-linked spermatid-expressed genes have become amplified in order to maintain a steady expression in the face of the enhancement of Sly-mediated repression and in a sense constitute a “collateral damage” arising from the conflict between Sly and Slx/Slxl1 that we unravel here . Interestingly , the Speer gene cluster is one of the autosomal gene families that have experienced the largest rodent-specific expansions [36] and is also repressed by SLY . Slx/Slxl1 and Sly competition may therefore have led to the amplification of reproductive genes outside the sex chromosomes as well as on them . F1 hybrid sterile males produced by asymmetric crosses between M . m . musculus and M . m . domesticus display sperm differentiation defects and wide-spread overexpression of X-encoded spermiogenic genes [37] . Intriguingly , this only occurs in males with a M . m . musculus X chromosome and M . m . domesticus Y and autosomal chromosomes [38] . These males have an excess of Slx/Slxl1 copies compared to Sly copies , since the M . m . domesticus X and Y chromosomes carry ∼40 to 60 copies of Slx/Slxl1 and Sly , while XY encoded Sycp3-related genes have been more amplified in M . m . musculus , with >100 of Slx/Slxl1 and Sly on the X and Y [18] , [20] . Our data show that a balance between Slx/Slxl1 and Sly expression exists in wild-type populations and that disruption of this balance can cause male infertility . In light of these data , we propose that deficiency in the number of Sly copies compared to Slx/Slxl1 copies contributes to F1 male hybrid sterility ( see Figure 6 ) in some of these crosses . This would explain the observed over-expression of X-encoded spermiogenic genes observed in some F1 hybrid males [37] and subsequent sperm differentiation defects and infertility . The observation that F1 males born from the reciprocal cross domesticus x musculus are reproductively normal [39] does not necessarily challenge this model . These males have an excess of Sly copies compared to Slx/Slxl1 copies and , according to our model , could be considered as Slx/Slxl1-deficient mice and thus display some spermiogenic defects . This however depends critically on the mechanism of the antagonistic effects of SLY and SLX/SLXL1 in the cytoplasm , and on the threshold of copy number imbalance required to trigger abnormal spermatogenesis and/or sex ratio skewing . Given that autosomal genes will be selected to maintain a balanced sex ratio , the Slx/Sly conflict may well be “buffered” to some extent by epistatic interactions with autosomal genes . We have observed that mice with a partial knockdown of Slx/Slxl1 ( shSLX1 or shSLX2 ) have comparatively minor spermiogenic defects compared to mice with a severe knock-down ( shSLX1/2 ) [28] . We also note that laboratory strain X chromosomes ( including MF1 mice which were used in the present study ) are predominantly derived from a domesticus background [40] , [41] , yet are paired in these strains with a musculus Y chromosome YRIII [42] . Thus laboratory strains are intrinsically comparable to the reciprocal cross . Our shSLX models therefore involve skewing the balance of SLX/SLXL1 and SLY even further , to pathogenic effect ( see Figure 6 ) . In this light it is intriguing that WT MF1 males have lower XY gene transcription than Slx/y deficient males: might this reflect the fact that laboratory strains are inherently “overdosed” for Sly relative to Slx/Slxl1 by virtue of their hybrid origin ? Male hybrid sterility is a complex trait involving several X-linked loci ( as demonstrated by the mapping of several quantitative trait loci – QTL – on the X chromosome [38] , [43] , [44] ) as well as autosomal factors [45] , [46] ) . It is worth noting that among the four non-overlapping X-chromosome-linked QTL associated with abnormal spermheads and hybrid sterility , one encompasses Slx ( 0–37 . 1 Mb ) , the other , Slxl1 ( 47 . 9–81 . 8 Mb ) [38] . Interestingly , it has been shown that one of the autosomal loci linked to hybrid sterility , Prdm9 , encodes a histone H3 lysine 4 methyltransferase involved in the silencing of the sex chromosomes during meiosis ( Meiotic Sex Chromosome Inactivation ) . It therefore epigenetically represses multiple X-chromosome loci , some of which part of the hybrid sterility gene network , and epistatic interactions between Prdm9 and multiple X and autosomal loci have been shown to cause asymmetric hybrid male sterility associated with a disruption of MSCI and thus a de-repression of the X chromosome [43] , [46] . However , Prdm9 does not appear to be involved in the X-chromosome up-regulation and sterility observed in F1 hybrid males studied by Good et al . [37] . Taken together , the genetic basis of reproductive isolation in mice is complex , and disruption of the transcriptional regulation of the X seems to contribute to the evolution of hybrid male sterility . The antagonistic effects of Slx/Slxl1 and Sly at the transcriptional and phenotypic level , in particular the effects on postmeiotic XY gene regulation , may therefore be among the important elements contributing to the evolution of hybrid sterility between mouse species . The production of F1 males with a transgene-derived increased Sly expression or with a knockdown of Slx/Slxl1 expression should help address this question . In conclusion , we have demonstrated that the mouse X and Y chromosomes are involved in an intragenomic conflict that is regulated by the multicopy genes Slx/Slxl1 and Sly . SLX/SLXL1 and SLY proteins compete during sperm differentiation , and notably have opposite effects on the regulation of sex chromosome gene expression . Disruption of Slx/y balance causes sex ratio distortion , sperm differentiation defects and male infertility . To the best of our knowledge , our work is the first characterization of a conflict over sex chromosome transmission in mammals and provides further evidence to support the hypothesis that intragenomic conflicts can have major consequences on gene regulation , genome structure and speciation .
shSLY ( aka sh367 ) , shSLX1 and shSLX1/2 males were produced and maintained as described before [19] , [28] . To produce shSLX1shSLY and shSLX1/2shSLY double transgenic mice , shSLX1 females were mated to shSLY or to shSLYshSLX2 transgenic males . Double transgenic females were then mated to MF1 XYRIII males ( see [19] ) to maintain the stock , since shSLY males are subfertile and give progeny only rarely . Two-month-old males single or double transgenic for sh367 ( shSLY ) , shSLX1 or shSLX1/2 transgenes , as well as their non-transgenic siblings ( WT ) were processed for all the analyses presented here . Animal procedures were in accordance with the United Kingdom Animal Scientific Procedures Act 1986 and were subject to local ethical review . Fractions enriched in round spermatids ( >90% ) were obtained from the above described transgenic and control ( WT ) males as described previously [19] . Each sample has been purified from a pool of testes obtained from 2 to 5 males . The coding sequence of mouse Dkkl1 and Slx cDNA were amplified by PCR and cloned into a C-terminal Myc-tagged pCMV vector; the coding sequence of mouse Slx , Slxl1 , Sly1 and Sly2 cDNA were amplified by PCR and cloned into a N-terminal Flag tagged pCMV vector using EcoRI and NotI restriction sites ( see Table S1 for a full list of primers ) . Co-transfections of HEK293 or COS cells were performed in 6-well plates using 1 . 5 µg of each DNA and 5 µl of Lifofectamine ( Invitrogen ) following the manufacturer's instructions . Proteins were extracted 24 hours post transfection in 200 µl of Lysis buffer ( 25 mM NaCl , 10 mM Tris-HCl , 5 mM EDTA , 0 . 1%NP-40 ) and immunoprecipitated as described below . Nuclear and cytoplasmic protein extracts were obtained as follow . The powder obtained from two adult testes crushed on dry ice was homogenized in a glass pestle with 1 mL of lysis buffer ( 0 . 6 M Sucrose , 10 mM Hepes pH 7 . 7 , 25 mM KCl , 2 mM EDTA , 0 . 5 mM EGTA and protease inhibitors ) . After the addition of 0 . 2% NP40 , the lysate was centrifuged for 15 minutes at 800 g . The supernatant corresponded to the cytoplasmic fraction . The pellet was washed twice with 1 mL of lysis buffer and then resuspended in 100 µl of nuclear protein extraction buffer ( 20 mM Hepes pH 7 . 7 , 1 . 5 mM MgCl2 , 0 . 2 mM EDTA , 25% glycerol and protease inhibitors ) plus 10 µl of 4 M NaCl . After 30 minutes of homogenization at 4°C , the samples were centrifuged for 30 minutes at 11000 g; the supernatant corresponded to the nuclear protein extract . A pellet of ∼1×107 round spermatids was extracted following the same protocol using 250 µl of lysis buffer and 50 µl of nuclear protein extraction buffer . Whole testicular protein extraction was performed as described previously [19] . For immunoprecipitation experiments , proteins extracted from transfected cells were first pre-cleared with protein A/G sepharose for 1 hour at 4°C . They were then incubated overnight with Protein G- or Protein A- sepharose which had been covalently bound to MYC ( Santa Cruz Biotechnology ) or FLAG ( Sigma ) antibody beforehand ( see [21] for a detailed protocol ) . Western blot experiments were performed as described previously [19] . Membranes were incubated overnight with anti-SLX/SLXL1 [28] diluted at 1/3000 , anti-SLY1 [21] at 1/3000 , anti–β-actin ( Sigma ) at 1/50000 , or anti-LAMIN B1 ( Santa Cruz Biotechnology ) at 1/1000 , anti-FLAG ( Sigma ) at 1/1000 , or anti-MYC ( Santa Cruz Biotechnology ) at 1/1000 . Detection by chemiluminescence was carried out after incubation with the corresponding secondary antibody coupled to peroxidase , as described by the manufacturer ( Millipore ) . Immunofluorescence experiments were performed on testis material fixed in 4% buffered paraformaldehyde and sectioned as described before [25] . DAPI ( 4′ , 6-diamidino-2-phenylindole ) was used to stain nuclei ( Vectashield DAPI , Vectorlab ) . Alexa Fluor 594-conjugated peanut agglutinin lectin ( Invitrogen ) , which stains the developing acrosome of spermatids , was used to stage the testis tubules . For the analysis of apoptotic elongating spermatids and delayed elongating spermatids , approximately 150 tubules were counted per individual ( 4 to 6 individuals per genotype ) . The percentage of tubules containing apoptotic elongating spermatids was determined on testis sections fluorescently stained using an in situ cell death detection kit ( TUNEL , terminal deoxynucleotidyltransferase dUTP nick end labeling ) as described by the manufacturer ( Roche Diagnostics , Indianapolis , IN ) . The percentage of tubules containing delayed elongating spermatids ( i . e . stage I to VIII tubules containing elongating spermatids ) was assessed on testis sections fluorescently stained by H4K12Ac antibody ( Millipore , Bedford , MA ) , a known marker of stage 9–12 elongating spermatids . Antibody detection was performed on surface-spread testicular cells following a protocol described previously [19] adapted from Barlow et al . [47] . Incubation with the primary antibody ( anti-SLY1 [17] , anti-SLX/SLXL1 [28] or anti-H3K9me3 [Upstate] diluted 1/100 ) was carried out over-night in a humid chamber at 37°C . DNA-FISH , then chromosome painting were performed after antibody detection as described previously [48] . Speer DNA-FISH was carried out using mouse BACs RP23-212A20 and RP24-310N20 ( CHORI ) . As a control for specificity ( see Figure S1 ) , SLX/SLXL1 antibody was preabsorbed with 8 mg of SLX immunogenic peptide or with 8 mg of a noncompeting peptide ( SLY peptide ) . For the quantification of H3K9me3 signal over the PMSC , the chromocenter domain was defined using the corresponding black and white DAPI picture . Then , H3K9me3 signal outside this chromocenter domain was measured and normalized to that of H3K9me3 signal over the chromocenter for each cell ( 100 cells per genotype ) , using Metamorph and ImageJ ( See Figure S4 ) . Slides corresponding to 3 individuals per genotype were coded and randomized before the analysis; the analysis was therefore carried out blind as to genotype . For the quantification of spermhead abnormalities , sperm collected from the initial caput epididymis were suspended in phosphate-buffered saline . The suspension was smeared on slides ( two slides per individual ) and fixed in 3∶1 methanol∶acetic acid . The slides were then dipped in 0 . 4% Photoflo for 2 min , air dried and stained on a plate heated at 60°C with one drop of 50% silver nitrate mixed with one drop of 2% gelatin ( Sigma ) . The slides were coded and randomized . Sperm scoring was carried out ‘blind’ as to genotype ( 4 to 6 individuals per genotype ) and 100 sperm per slide were classified into 6 categories on the basis of the type and severity of abnormalities observed , using criteria described by Yamauchi et al . [49] and in Figure S6 . In the text and figures , spermheads from category N were termed “normal”; category 1S , “slightly abnormal”; category 2S , “slightly flattened”; category 3G , “abnormally thin”; category 4G , “grossly flattened” and categories 5G to 8G were pooled and named “other gross abnormalities” ( cf . Figure S6 ) . To assess fertility and obtain sex ratio data from the offspring , five males of each genotype were mated with MF1 WT females over a period of six months . Real-time Reverse Transcription-Polymerase Chain Reaction ( RT-PCR ) and microarray analyses were performed as previously described on RNA extracted from 2-month old testis or from round spermatids obtained after elutriation [19] ( cf . Table S1 for a list of the primers used in this study ) . Real-time RT-PCR experiments were performed in parallel for all the genotypes described in this study , with between 3 to 5 individuals per genotype . For the microarray analysis , three shSLX1 , three shSLY , three shSLX1shSLY and four wild type spermatid batches were analyzed ( Illumina BeadChip , mouse whole-genome array , v2 ) . These data thus include and extend our previously-reported results for shSLX1 round spermatids and for shSLY round spermatids in previous analyses [19] , [28] , which collectively used two shSLY , two shSLX1 and four WT spermatid batches . Data normalization and calculation of FDR-adjusted p values was carried out in BeadStudio ( Illumina ) as previously described [19] , [28] . The full data set has been uploaded to GEO , accession number GSE39109 . For comparisons of the incidence of sperm head abnormalities , differences between genotypes were assessed by ANOVA after angular transformation of percentages , using the General Linear Models ANOVA provided by NCSS statistical data analysis software . The same test was applied to the frequency of abnormal head-tail connections , TUNEL positive elongating spermatids , delayed elongating spermatids ( assessed by H4K12Ac staining ) and H3K9me3 quantification . Student's t test was used to compare the data obtained for fecundity , sperm number and real-time PCR ( performed on the ΔCt values ) . A Chi-square test was used for sex ratio data . Microarray results were analyzed as described in Figure S3 and Figure S5 .
|
Both copies of a gene have normally an equal chance of being inherited; however , some genes can act “selfishly” to be transmitted to >50% of offspring: a phenomenon known as transmission distortion . Distorting genes on the X or Y chromosome leads to an excess of female/male offspring respectively . This then sets up a “genomic conflict” ( arms race ) between the sex chromosomes that can radically affect their gene content . Male mice that have lost part of their Y produce >50% female offspring and show over-activation of multiple genes on the X , providing strong circumstantial evidence for distortion . Here , we demonstrate the existence of a genomic conflict regulated by the genes Slx/Slxl1 and Sly , present in ∼50 to 100 copies on the X and Y chromosomes respectively . SLX/SLXL1 and SLY proteins have antagonistic effects on sex chromosome expression in developing sperm and skew the offspring sex-ratio in favor of females/males . Interestingly , while deficiency of either gene alone leads to severe fertility problems , fertility is improved when both genes are deficient . We believe that the conflict in which Slx/Slxl1 and Sly are involved led to the amplification of X and Y genes and may have played an important role in mouse speciation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"genome",
"evolution",
"evolutionary",
"biology",
"chromosome",
"biology",
"gene",
"expression",
"genetics",
"biology",
"genomics",
"genomic",
"evolution",
"genetics",
"and",
"genomics",
"cell",
"differentiation",
"gene",
"function"
] |
2012
|
A Genetic Basis for a Postmeiotic X Versus Y Chromosome Intragenomic Conflict in the Mouse
|
Bacterial virulence is a multifaceted trait where the interactions between pathogen and host factors affect the severity and outcome of the infection . Toxin secretion is central to the biology of many bacterial pathogens and is widely accepted as playing a crucial role in disease pathology . To understand the relationship between toxicity and bacterial virulence in greater depth , we studied two sequenced collections of the major human pathogen Staphylococcus aureus and found an unexpected inverse correlation between bacterial toxicity and disease severity . By applying a functional genomics approach , we identified several novel toxicity-affecting loci responsible for the wide range in toxic phenotypes observed within these collections . To understand the apparent higher propensity of low toxicity isolates to cause bacteraemia , we performed several functional assays , and our findings suggest that within-host fitness differences between high- and low-toxicity isolates in human serum is a contributing factor . As invasive infections , such as bacteraemia , limit the opportunities for onward transmission , highly toxic strains could gain an additional between-host fitness advantage , potentially contributing to the maintenance of toxicity at the population level . Our results clearly demonstrate how evolutionary trade-offs between toxicity , relative fitness , and transmissibility are critical for understanding the multifaceted nature of bacterial virulence .
The development of effective , long-term control strategies against microbial pathogens crucially relies on a thorough understanding of the many factors that contribute to the evolution and maintenance of enhanced virulence . Bacterial toxins are well established as playing a key role in virulence [1–4] . They release nutrients for bacterial growth and facilitate intra- and interhost transmission by destroying local tissue and subverting host immune processes [5–7] . This has led to the general presumption that elevated toxicity is positively associated with enhanced disease severity in bacterial infections [5 , 8 , 9] . However , the expression of toxins is readily switched off in vitro in response to the selection imposed by the energetically costly nature of their production [10 , 11] . Observational studies in S . aureus suggest that this can also occur in vivo , indicating that the relationship between toxicity and disease severity is more complex than initially appreciated [12–17] . S . aureus is a major human pathogen and a global healthcare issue . It is considered opportunistic as it asymptomatically colonises its host but can occasionally cause diseases that range in severity from relatively minor skin and soft tissue infections ( SSTI ) to life-threatening cases of pneumonia and bacteraemia [9] . Toxicity has been accepted as playing a key role in the success of lineages such as USA300 [18] and ST93 [19] , in which it has been suggested to increase transmissibility . Toxicity is also widely accepted as playing a significant role in the virulence of S . aureus , where many studies comparing high- and low-toxicity isolates in animal models of sepsis show that highly toxic isolates cause more severe disease symptoms [20–24] . However , it has recently been shown that S . aureus isolates from humans with invasive diseases , such as bacteraemia and pneumonia , are often impaired in their ability to express toxins ( often referred to as Agr dysfunction ) [12 , 13] . A strength of the collections of isolates considered in these studies is that they represent all of the isolates presenting in a given geographical region over a defined period of time . Unfortunately , this type of sampling brings with it a limitation , as such collections consequently also contain a wide range of genetic backgrounds and antibiotic resistance profiles that might confound any potential associations between toxicity and disease . In this study , we applied a robust functional genomics approach to two collections of S . aureus isolates , with significant depth and breadth , where we have controlled for genetic background by sampling within specific clinically important lineages , including USA300 . Our approach not only allowed us to make observations that challenge our understanding of the role of toxicity in the establishment of severe , invasive disease but also to identify the genetic polymorphisms involved . By using a combination of functional assays , we further identified bacterial fitness in human serum as an important factor that could limit the ability of highly toxic isolates to cause bacteraemia and would explain the observed negative relationship between toxicity and disease severity . The power of genomics to study past events is clear , and here we demonstrate its potential to also help us understand fundamental aspects of bacterial pathogenicity and their role in invasive disease .
The single-patient collection included serial asymptomatic nasal carriage isolates over a 12-mo period , as well as bloodstream isolates after bacteraemia had developed at month 15 [25] . At each time point , 12 individual colonies were isolated from the primary plates for each swab . These isolates all belong to the ST15 lineage and contain genes for 12 of the 13 known cytolytic toxins secreted by S . aureus ( i . e . , alpha , beta , gamma , delta , LukAB , LukED , PSMα1 , α2 , α3 , α4 , β1 , β2 , but not LukSF [a . k . a . Panton-Valentine-leukocidin or PVL] ) . To quantify the gross cytolytic activity of each isolate , we used an immortalised T/B hybridoma cell line ( T2 ) , which is susceptible to 10 of the 12 cytolytic toxins present in this collection ( not LukAB or LukED ) ( S1 Fig ) . Despite some sequence variability across the 12 isolates from each time point and between the different time points [25] , no diversity in toxicity was observed for the early nasal culture isolates ( Fig 1A ) . At month 12 , however , there was a significant drop in toxicity for all 12 nasal carriage colonies , shortly after which the study participant developed bacteraemia ( month 15 ) . The bacteria isolated from the patient’s bloodstream also showed significantly reduced toxicity compared to those from the earlier time points ( Fig 1A ) , demonstrating an apparent inverse correlation between toxicity and disease , albeit in a sample size of only one patient . To ensure this effect was not a consequence of the lack of sensitivity of the T2 cell line to LukAB and ED , we quantified the ability of a subset of six high- and six low-toxicity isolates ( as determined using the T2 cells ) to lyse freshly harvested human neutrophils , which are sensitive to LukAB and LukED . We found that the isolates that were unable to lyse the T2 cells were also unable to lyse the neutrophils ( S1 Fig ) , suggesting that LukAB and LukED were not expressed by these low-toxicity isolates , and that the T2 cell line provides a robust measure of gross cytolytic activity for this collection of isolates . Having demonstrated a negative correlation between toxicity and disease for a single patient , we sought to extend this to multiple patients by focussing on a collection of 134 USA300 MRSA isolates . As the USA300 lineage is known to contain the genes encoding the LukSF ( i . e . , PVL ) toxin we included a human monocytic cell line , THP-1 , which is sensitive to the action of this toxin [27] . In this set of isolates , we again found that the bacteraemic isolates were significantly less toxic than either the carriage or the SSTI groups ( Fig 1B and 1C , ANOVA: p < 0 . 01 for both T2 cells and THP-1 cells ) . With these two collections of isolates , representing both broad sampling across patients and deep sampling within a patient , we demonstrate a significant negative association between toxicity and disease severity . The single-patient collection contained only a small number of genetic differences between the isolates , and as such we were readily able to identify the genetic basis of the change in toxicity . We found that the late nasal carriage and blood culture isolates had a premature stop-codon in an araC-like transcriptional regulator gene rsp , which encodes a protein that has previously been shown to regulate biofilm formation by this pathogen [28] . Using a transposon insertion in the rsp gene ( NE1304 ) from the Nebraska library [29] , we found a significant reduction in toxicity compared to the wild-type strain ( Fig 2 ) , thus verifying this gene’s toxicity-regulating function . Interestingly , mutating the rsp gene with a transposon in the USA300 ( ST8 ) background ( Fig 2 ) did not have as significant an effect on toxicity as the stop-codon in the ST15 background ( Fig 1A ) , presumably as a consequence of differences in the genetic background of these bacterial clones . To identify the genetic polymorphisms responsible for the changes in toxicity of the USA300 isolates , we adopted a functional genomics approach using genome-wide association studies ( GWAS ) to identify candidate polymorphisms associated with toxicity ( S2 Table ) . As GWAS is prone to high false positive rates , we sought to verify the effect of the associated loci using transposon insertions in each locus that was available from the Nebraska library ( the full list of significantly associated polymorphisms can be found in S2 Table ) . We identified five novel toxicity-affecting loci ( Fig 2 ) : ftsK , clpC , sucD , rpsA , plus a hypothetical gene with no known activity or homology to other proteins ( SAUSA300_0750 ) . Based on amino acid homology , the FtsK protein is believed to be a DNA translocase , where polymorphisms have also been associated with changes in the toxicity of another globally successful MRSA clone , ST239 [30] . FtsK also shares significant structural similarities to transporters , suggesting that this protein may be directly involved in the secretion of toxins from the bacterial cell , rather than DNA translocation . The ClpC protein is a chaperone and has been shown previously to affect the expression of a large number of proteins , including several regulators of toxin expression [31] . The role of the proteins encoded by sucD ( succinyl-CoA synthetase subunit alpha ) and rpsA ( 30S ribosomal protein S1 ) in toxin expression is less clear , although any changes to the metabolism of a cell are likely to have significant downstream effects on gene expression [32] . Further work is underway to elucidate the molecular detail of how these proteins affect toxicity . With mortality rates as high as 20% for S . aureus bacteraemia [40] , understanding how these types of infections develop is of significant clinical importance . However , as the bacteria rarely transmit beyond this point , they can be effectively thought of as transmission dead ends . We therefore sought to explore the possible evolutionary effect of the observed association of low-toxicity isolates with invasive infections using a simple mathematical model ( see Methods ) . We considered two competing strains of S . aureus that differed in their level of toxicity , where higher toxicity was assumed to be positively correlated with transmissibility and progression from carriage to SSTI , based on the fact that we rarely find low-toxicity isolates amongst our carriage or SSTI populations , and that low-toxicity mutants have been rarely found to transmit amongst healthy populations [17] . But higher toxicity also resulted in faster treatment . With this , we examined what effect the strains’ relative propensity to cause bacteraemia ( σ ) had on their competitive fitness . Assuming no differences ( σl = σh , Fig 6A and 6B and S3 Fig ) , we find that the higher clearance rate of the more toxic strain offsets its transmission advantage , leading to its exclusion and dominance of the less toxic strain ( shown separately for carriage and SSTI and bacteraemia in Fig 6A and 6B , respectively ) . In contrast , considering the clinically observed negative association of toxicity with bacteraemia ( σl > σh ) results in the more toxic strain gaining a competitive advantage at the population level ( carriage and SSTI , Fig 6C , S3 Fig and S4 Fig ) , whereas the less toxic strain maintains its elevated frequency during bacteraemia ( Fig 6D , S3 Fig and S4 Fig ) , in line with clinical observations . This suggests that the reduced opportunity for transmission due to bacteraemia could partly compensate for the toxicity-driven trade-off between transmissibility and treatment rates and thus contribute to the maintenance and circulation of highly toxic strains within the population .
For microbial pathogens , many factors contribute to their success , but for an opportunistic pathogen that can either reside asymptomatically or cause symptomatic infections ranging from superficial to life-threatening invasive disease , the definition of success becomes increasingly complex . The damage-response framework outlines the necessary holistic approach we need to take when considering this , where both the level of virulence expressed by the pathogen and the response of the host is critical to the clinical outcome [33] . For S . aureus , the vast majority of infections resolve without clinical intervention , and severe infections ( e . g . , bacteraemia or pneumonia ) are generally limited to those with compromised health [9] . With many genes encoding cytolytic toxins , and highly toxic clones disseminating worldwide , it is understandable that toxicity and virulence have long been considered directly linked and key to its success . However , here we show this relationship to be quite complex . The boom in genomic sequence data for microbial pathogens has allowed us to study past genetic events in great detail , tracing epidemics and studying how bacterial genomes evolve [25 , 41–44] . It is , however , only recently that we have been able to successfully use such genomic data to study the behaviour of pathogens and use functional genomics approaches to understand why and how specific traits evolve [30 , 45–49] . By studying two large collections of isolates , we demonstrate that bacteraemic isolates are significantly less toxic than those isolated from carriage or from SSTIs . With the genome sequence of each isolate available to us , we were able to identify the polymorphisms responsible for the observed changes in toxicity and in doing so have identified six novel toxicity-affecting loci for this pathogen . A molecular dissection of each locus is currently underway to determine how they affect this trait , but this work clearly demonstrates the power of functional genomics for studying bacteria . To understand why low toxicity isolates have a higher propensity to cause bacteraemia , we developed and tested several hypotheses . While the health of a patient is a feature in their susceptibility to bacteraemia , we were unable to find evidence to suggest that this would increase their propensity to develop bacteraemia with a low rather than a highly toxic isolate . Furthermore , we found no evidence suggesting that either cell-invasiveness , NET formation , protease activity , antimicrobial peptide resistance , or biofilm formation play a role . Instead , we found that the presence of serum , which simultaneously increases toxin expression while neutralising their activity , reduces the relative fitness of the highly toxic isolates . Given the extreme bottleneck that the establishment of a bloodstream infection represents to bacteria , we believe this explains our finding that low-toxicity isolates are a more common cause of this type of invasive disease . By use of a mathematical model , we also showed that this increased propensity for low-toxicity isolates to cause bacteraemia , alongside the dead end nature of such infections , could potentially contribute to the maintenance of high levels of toxicity at a population level , as evidenced by the global prevalence of highly toxic clones , such as USA300 and ST93 . To understand the relative efficiency of GWAS to identify novel toxicity-affecting loci , we need to compare it to a similar experiment that used a random approach . Fey et al . , who created the transposon library used here , screened their 1952 Tn mutants for haemolytic activity using rabbit blood agar plates and identified 71 mutants with a change in this phenotype [29] . Compare this hit rate of 3 . 6% with ours of 6 . 7% and the relative efficiency of this GWAS approach is apparent . This efficiency is , however , significantly affected by the size and relatedness of the isolates within the collection; as in a previous GWAS study , we used a more closely related collection of isolates and had a hit rate of 30% [30] . Such issues need to be considered when designing GWAS experiments for bacteria . This work highlights a clear limitation to existing animal models of infection for pathogens like S . aureus . Studies published by us and many other groups have shown that in sepsis models , highly toxic isolates cause a more severe infection in mice than less toxic isolates [20–24] , which would seem contrary to our findings here in humans . However , a major difference between these two systems needs to be considered . For humans , studies have demonstrated that bacteraemia represents an extreme bottleneck , with only a small number of cells being sufficient to cause disease [25 , 50] . For mice , on the other hand , 107–108 colony-forming units need to be injected directly into the tail vein to get reproducible infections , which overrides the rigours bacteria need to go through to cause bacteraemia naturally in human . Furthermore , subtle differences in relative fitness , which we suggest contribute to our clinical observations , would not be evident in a mouse overwhelmed with the introduction of so many bacteria cells directly into its bloodstream . We would therefore urge caution when interpreting the results from such experiments in relation to human disease . It is interesting to note that no difference in toxicity was observed between the carriage isolates and those causing SSTIs , and that , on average , both groups were highly toxic . Many animal models have been used to demonstrate that the size of a cutaneous lesion is greatly affected by the levels of toxins secreted by the infecting organism [5] . While we have no clinical data to compare toxicity levels and lesion size for these isolates , that we rarely find low-toxicity isolates causing SSTIs supports the findings from animal models that toxicity correlates with the ability to cause such lesions . It is also intriguing to consider what selective forces maintain such a high level of toxicity at a population level . For symptomatic transmission , the toxin-induced production of highly transmissible pus has obvious benefits . However for asymptomatic transmission , this is less obvious . Low toxicity isolates can survive in the nose , as indicated by the single patient collection where the low toxicity isolates survived there for 3 mo before causing bacteraemia . However , perhaps over a longer term , the ability of highly toxic bacteria to resist the effects of nasal-associated immunity by killing host immune cells , and the use of some toxins ( e . g . , the phenol soluble modulins , PSMs ) to interfere with competing genera of bacteria , is key to its maintained selection . For S . aureus , an opportunistic pathogen , it is clear that virulence is multifaceted . On the one hand , the prevalence of highly toxic clones globally and the role of specific toxins in causing highly transmissible pus-filled SSTI lesions suggests that toxins offer a selective advantage . On the other hand , it appears that offsetting toxicity for short-term enhanced fitness is associated with increased virulence , which may paradoxically select for the maintenance of higher levels of toxicity at a population level . Although at a superficial level these seem contradictory , it is clear that they are critical aspects of this pathogen’s success . With the movement of genome sequencing into routine clinical practice and the drive towards personalised medicine , we need to define these complex interactions and bring the biology of the pathogen into greater consideration in clinical settings .
Whole blood was obtained from healthy volunteers; ethical permission for all donations was obtained from local research ethics committee ( School of Pharmacy & Pharmaceutical Sciences Ethics Committee ) and all participants gave written informed consent . A list of the bacterial strains used can be found in S1 Table . For the toxicity assays , the S . aureus isolates were grown overnight in 5 mL of Tryptic-Soy Broth ( TSB ) in a 30 mL glass tube . This overnight culture was used to inoculate the toxicity-assay cultures at 1:1 , 000 dilution in fresh TSB and then grown for 18 h at 37°C in air with shaking ( 180 rpm ) . For transposon mutants , erythromycin ( 5 μg/mL ) was included in the growth medium . The toxin-containing supernatant for each isolate was harvested by centrifugation . Immortalised human T2 cells and monocyte macrophage THP-1 cell lines were used as described previously [51] . The THP-1 cell line was included for the USA300 collection as it is susceptible to the PVL [27] . Briefly , both cell lines were grown in individual 30 mL suspensions of RPMI-1640 , supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) , 1 μM L-glutamine , 200 units/mL penicillin , and 0 . 1 mg/mL streptomycin at 37°C in a humidified incubator with 5% CO2 . Cells were routinely viewed microscopically every 48–60 h and harvested by centrifugation at 1 , 000 rpm for 10 min at room temperature and resuspended to a final density of 1–1 . 2 x 106 cells/mL in tissue-grade phosphate buffered saline . This procedure typically yielded >95% viability of cells as determined by trypan blue exclusion and easyCyte flow cytometry . To monitor S . aureus toxicity , 20 μL of cells were incubated with 20 μL of bacterial supernatant and incubated for 12 min at 37°C . For the USA300 strains , supernatants were diluted to 30% of the original volume in TSB as these isolates were considerably more toxic than the single-patient isolates . Cell death was quantified using easyCyte flow cytometry using the Guava viability stain according to manufacturer’s instructions . Experiments were done in triplicate , and error bars indicate the average ± the 95% confidence interval of multiple independent experiments . The identification of genetic variation in all the clinical isolates studied has previously been described [25 , 26] with the exception of the 36 bacteraemic USA300 isolates . These were sequenced in an identical manner to the others; namely , genomic DNA was extracted using the QIAamp DNA Mini Kit ( Qiagen ) , and unique index-tagged libraries were generated . Whole-genome sequencing was carried out using the Illumina HiSeq2000 with 100-base paired-end reads . Paired-end reads were mapped against the core chromosome of the ST8 USA300 reference genome sequence FPR3757 ( accession NC_02952 ) [52] . SNPs and indels were identified as described previously [53] . ENA accession numbers are listed in S1 Table . We conducted a quantitative association study on a set of 134 USA300 isolates to identify SNPs that were significantly associated with toxicity , using the PLINK software package ( http://pngu . mgh . harvard . edu/purcell/plink/ ) [54] . These and a description of the loci are listed in S2 Table . This was estimated using PhyML with an HKY85 substitution model , empirical nucleotide usage , no rate heterogeneity , and no invariant sites . The percentage toxicity range was divided by three , where the most toxic isolates were labelled red , the midtoxicity isolates labelled orange , and the least toxic isolates labelled green . Bacterial invasion of EA . Hy926 endothelial cells were performed as described previously with minor modifications [55] . Endothelial cells were cultured in Dulbecco’s modified Eagles’ medium ( DMEM ) supplemented with 10% FBS and 2 mM L-glutamine at 37°C in a humidified incubator with 5% CO2 . Cells were liberated using trypsin-EDTA solution , resuspended in culture medium , and aliquoted into 24-well tissue culture tissue plates and grown to >95% confluence . Cells were washed twice in tissue-grade PBS , and 450 μL of fresh DMEM was added . 50 μL of washed S . aureus ( 1 x 107 CFU/mL ) was added to the wells and incubated for 1 h at 37°C . Following incubation , the medium was aspirated and wells gently washed once in PBS and replaced with DMEM supplemented with 200 μg/mL gentamicin and incubated at 37°C for a further 60 min . Cells were subsequently lysed by the addition of 500 μL of Triton X-100 , and bacterial CFU were enumerated by serial dilution of endothelial cell lysates and plating onto TSA plates and incubated at 37°C overnight . Experiments were performed in duplicate three times , and the error represents the 95% confidence interval . Whole blood was obtained from healthy volunteers; ethical permission for all donations was obtained from a local research ethics committee ( School of Pharmacy & Pharmaceutical Sciences Ethics Committee ) , and all participants gave written informed consent . Human neutrophils were isolated as previously described [56] . Cell-free supernatants from bacterial culture were diluted to 30% in warm Krebs buffer and then diluted 1:1 with prewarmed neutrophils ( 106 neutrophils/mL in Krebs ) and incubated at 37°C for 12 min . Cells were pelleted , and NET formation was quantified by measuring DNA content in the supernatant with Sytox Green and a DNA standard curve . Any signal from the bacterial culture was measured and subtracted from these values . For the neutrophil lysis assay , 20 μL of purified neutrophils was incubated with 20 μL of 10% bacterial supernatant for 15 min , and cell viability was assayed using Guava viability reagent and Guava flow cytometry . A modified tryptic soy agar medium was made with 10% skim milk . 50 μL of bacterial supernatant harvested from overnight cultivation was inoculated into 1 cm diameter wells perforated into the agar medium and incubated for 18 h at 37°C . The digested substrate , as a result of proteolytic activity , was observed as clear areas surrounding the wells , was measured . Protease assay were done in duplicate , three times , and error represents the 95% confidence interval . Purified human neutrophil defensin-1 ( hNP-1 ) was purchased from AnaSpec Incorporated ( California , USA ) . The hNP-1 susceptibility assay was performed in 1% BHI with the addition of 10 mM potassium phosphate buffer as described previously [57] . A final inoculum of 105 CFU , with a peptide concentration of 5 μg/mL , was employed and incubated for 2 h at 37°C . Final bacterial concentration was evaluated by serial plating onto TSA plates and data represented as mean ( ± SD ) percent survival CFU/mL . Semiquantitative measurements of biofilm formation on 96-well polystyrene plates was determined based on the method of Ziebuhr et al [58] . Overnight bacteria grown in TSB were diluted 1:40 into 100 μL TSB containing 1% glucose and grown for 24 h at 37°C . Following 24-h growth , plates were washed vigorously five times in PBS , dried and stained with 150 μL of 1% crystal violet for 30 min at room temperature . Following five washes of PBS , wells were resuspended in 200 μL of 7% acetic acid , and optical density at 595 nm was recorded using a plate reader . S . aureus isolates were grown overnight in BHI broth to an OD600 of 2 . 0 to ensure that all cells are in a similar physiological state at the start of the experiment . Competitions were established in TSB with and without 5% ( vol/vol ) freshly drawn human serum . The competition medium was inoculated with 104 CFU/mL of the marker strain ( MSSA466 , which is tetracycline resistant ) and 103 CFU/mL of the test strain . Initial cell numbers were confirmed by plating . The bacteria were competed at 37°C in a shaking incubator ( 180 rpm ) for 24 h . Final cell numbers were enumerated by serial dilutions on TSA plates ( total cell count ) and TSA plates containing 2 μg/mL tetracycline ( marker strain count ) . The fitness of a strain was defined as a measure of the reproductive success of the population , which can be expressed as the natural logarithm of the ratio of the final and initial cell densities of the culture [59] . Each clinical strain was assayed once as each was considered a biological replicate indicative of its group . relative ( Darwinian ) fitness=ln ( A ( 1 ) /A ( 0 ) ) ln ( M ( 1 ) /M ( 0 ) ) where: A ( 0 ) , estimated density of test strain at time 0; M ( 0 ) , estimated density of marker strain at time 0; A ( 1 ) , estimated density of test strain at time 1 d; M ( 1 ) , estimated density of marker strain at time 1 d; ln , natural logarithm ( logarithm to the base e ) . We used a simple transmission model to examine the qualitative effect of toxicity-dependent probabilities of S . aureus to cause invasive disease ( see S4 Fig for a flow diagram ) . We considered two different strains , distinguished by their level of toxicity ( high , h , and low , l ) , and assumed that susceptible individuals , S , become colonised with strain i ( I = h , l ) upon contact with either colonised or infected individuals ( Ci and Ii ) with transmission rates βc and βi , respectively . Individuals transition from colonisation to infection ( i . e . , SSTI ) at a rate δi , from which they either recover ( at rate τi and ρi , accounting for both treatment and immune mediated clearance ) or go on to develop invasive diseases ( bacteraemia , Bi ) , with probability σi , which we assume does not contribute to transmission . The model can then be given by the following set of differential equations for the proportions of the population being susceptible , colonised , infected , or suffering from invasive disease: dSdt=μ-S ( λl+λh ) +τlIl+τhIh-μS dChdt=λhS+ϱhIh- ( μ+δh ) Ch-νCh dCldt=λlS+ϱlIl- ( μ+δl ) Cl+νCh dIhdt=δhCh-ϱhIh- ( μ+σh+τh ) Ih-νIh dIldt=δlCl-ϱlIl- ( μ+σl+τl ) Il+νIh dBidt=σiIi- ( μ+χ ) Bi , i=h , l with the force of infection of strain i , λi = βCCi + βiIi . , disease-induced mortality χ , and μ as the natural birth/death rate . For simplicity , we did not allow for co- or superinfections but considered within-host evolution whereby more toxic strains can mutate ( at a rate ν ) towards lower levels of toxicity . Within this system , the probability of colonised individuals to develop infections ( SSTI’s ) was assumed to be positively correlated with the strain’s degree of toxicity ( with δh > δl ) , as were the transmission and treatment rates of infected individuals ( i . e . , βh > βl , τh > τl ) . For illustration purposes only , we assumed that when the strains have equal probabilities to develop invasive disease ( i . e . , σ1 = σ2 ) , the less toxic strain has a higher fitness than the more toxic one . That is , we assumed that the less toxic strain is at an optimum , whereby toxicity-driven increases in transmissibility would be offset by higher clearance rates . Unless stated otherwise , we used the following parameter values: μ = 0 . 017 , βC = 0 . 05 , βl = 4 , βh = 4 . 4 , δl = 2 , δh = 2 . 2 , τl = 3 , τh = 3 . 3 , ρl = ρh = 10 , ν = 0 . 002 , χ = 5 .
|
Global efforts to counter the growing problem of antibiotic resistance and develop alternative treatment strategies rely on a fuller understanding of when and why opportunistic pathogens cause disease . Recent advances in DNA sequencing technologies have opened up new opportunities to study infectious organisms , yet identifying the genetic variants that explain differences in disease remains challenging . Here we aimed to understand the complex relationship between toxicity—a known risk factor for disease in many bacteria—and infection severity for the major human pathogen S . aureus . Against expectations , we found that the bacteria that caused the most severe disease were the least toxic strains . We were able to determine the mutations responsible for the differences in toxicity and identified a number of novel toxicity-affecting genes . We further discovered that bacterial fitness in human serum could explain the unexpected association of low-toxicity isolates with severe , invasive disease . Invasive S . aureus infections are usually considered a dead end for these bacteria , as these infections are rarely transmitted to another person . Here we show using a simple mathematical model that this might in fact favour transmission of highly toxic bacteria on a population level and thus contribute to their global success . Our work therefore highlights the complexity of bacterial infection and should aid in devising new treatment and control strategies against this important pathogen .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Evolutionary Trade-Offs Underlie the Multi-faceted Virulence of Staphylococcus aureus
|
Non-domiciliated intrusive triatomine vectors are responsible for a low but significant transmission of Trypanosoma cruzi to humans . Their control is a challenge as insecticide spraying is of limited usefulness , and alternative strategies need to be developed for a sustainable control . We performed a non-randomized controlled trial of an Ecohealth intervention based on window insect screens and community participation to reduce house infestation by Triatoma dimidiata in two rural villages in Yucatan , Mexico . Efficacy of the intervention was measured over a three years follow-up period and entomological indicators showed that the proportion of triatomines found inside houses was significantly reduced in houses with insect screens , which effectively kept more bugs on the outside of houses . Using a previously developed model linking entomological data to the prevalence of infection in human , we predicted that the intervention would lead to a 32% reduction in yearly incidence and in the prevalence of T . cruzi infection . The cost for the coverage of all the windows of a house was of comparable magnitude to what families currently spend on various domestic insecticide , and most screens were still in good conditions after three years . In conclusion , the Ecohealth approach proposed here is effective for the long-term and sustainable control of intrusive T . dimidiata vectors in the Yucatan peninsula , Mexico . This strategy may also be easily adapted to other intrusive triatomine species as well as other regions/countries with comparable eco-epidemiological settings , and would be an excellent component of a larger integrated program for the control of a variety of other vector-borne diseases , bringing additional benefits to the communities . Our results should encourage a further scaling-up of our implementation strategy in additional villages in the region .
Chagas disease is a vector-borne parasitic disease causing major morbidity and mortality in the Americas , with at least 6 million persons currently infected . The burden of Chagas disease is estimated at 29 million disability-adjusted life years ( DALYs ) and it leads to health care costs of $24 . 73 billion [1 , 2] . In Mexico , the Ministry of Health reports a few hundred cases every year [3] , but estimates suggest that there may be over one million infected persons [4] . In the state of Yucatan , seroprevalence surveys have reported that 1–4% of the population is seropositive for Trypanosoma cruzi , the agent of Chagas disease [5–7] . Throughout the Americas , Chagas disease is principally controlled by residual insecticide spraying to reduce house infestation by triatomine vectors [8 , 9] . However , the emergence of insecticide resistance [10] and the limited usefulness of insecticide spraying against intrusive vectors make such interventions clearly not sustainable for long-term vector control [9 , 11 , 12] and evidences the need for better and integrated vector control interventions [8] . For example , the Ecohealth approach ( ecosystem approach to health ) promotes interventions targeting the multiple determinants of disease transmission through transdisciplinary participatory research , integrating biological , ecological and social aspects of disease control [13–15] . Ecohealth strategies are emerging as more rational , sustainable , and cost-effective than vertically-organized and widespread empirical insecticide spraying , as they promote greater community responsibility and ownership of the intervention [16] . Triatoma dimidiata in the Yucatan peninsula , Mexico , is a good example of intrusive triatomines that transiently infest houses during the months of March-July [17–21] , and which is difficult to control with insecticide [11 , 20] . We previously identified some of the key determinants for house infestation by T . dimidiata in the region , which include house proximity to sylvatic areas , proximity to public lighting and the presence of domestic animals such as dogs and chicken , while housing quality or domestic practices have little relevance [22 , 23] . Based on qualitative research , we assessed community knowledge , attitudes and practices related to triatomine vectors and Chagas disease and vector control [24 , 25] . This information provided some clues to develop an Ecohealth intervention based on a community program of insect screen manufacture and installation for triatomine control in two villages [26] . Here , we further evaluated the efficacy of the intervention during three consecutive infestation seasons , based on entomological indicators .
The current study was performed in the rural villages of Teya ( 21 . 05°N , 89 . 07°W ) , Sudzal ( 20 . 87°N , 88 . 98°W ) , and Bokoba ( 21 . 01°N , 89 . 07°W ) , located about 15–20 km apart in the central part of the Yucatan state in southern Mexico [24] . Climate in the region is warm and humid , with an average annual temperature of 26°C and 115 cm of rainfall . The villages are surrounded by a mixture of secondary bush vegetation and agricultural land . There is a total of 702 , 509 and 570 houses in Teya , Sudzal ( main village ) and Bokoba , respectively , all of which have been georeferenced . The village of Sudzal also includes the communities of Tzalam , Chumbec and Kancabchén , which represent an additional 120 houses . The respective population is of about 2 , 000 inhabitants in both Teya and Bokoba , and 1 , 600 in Sudzal . All three villages are very similar and comparable in most relevant aspects , including triatomine infestation and T . cruzi seroprevalence in the population [7 , 22–24] . The study is a non-randomized controlled trial , in which vector control was implemented in two communities ( Teya and Sudzal ) , and control households were those of both communities which did not participate in the intervention . Indeed , enrolment of the households in the project was voluntary and community-based , as described below . Consequently , allocation of mosquito screens was not randomized . For years 2 and 3 of follow-up , households of a third community ( Bokoba ) , where no vector control was implemented , served as additional controls . As many individuals and organizations as possible were initially approached in each community to discuss the project , allowing them to express their potential interest and contribution , and leading to the identification of stakeholders in each community as well as providing key information for a situation analysis [22 , 24 , 25] . The intervention was then designed through a series of open meetings with the different stakeholders . These included the communities , local governments , local health centers , social workers , carpenters , and community leaders . As previously described [26] , different implementation strategies were discussed to identify the preferred process in each community ( Teya and Sudzal ) . Local carpenter workshops were identified by the communities and the local governments , to ensure better community dynamics and ownership of the intervention Two carpenter workshops were initially identified in Teya and one in Sudzal . However , one of the workshops in Teya withdrawed from the project shortly after initiating the intervention , due to lack of interest . Thus screens were manufactured and installed by one carpenter workshop in each village for most of the intervention . They arranged for home visits to each community member to take measurements of windows , delivery and installation of the screens . Since the large majority of households only had one bedroom , it was agreed that a single bedroom would be protected ( on average with 2 windows ) . Social workers from the municipal team ( Teya ) and from the health center ( Sudzal ) took responsibility to coordinate the activities related with the implementation of the interventions , following training by the research group . This included the organization of weekly community meetings at the city hall , coordinating the carpenters for the distribution of materials and supervising the installation of screens , and administrating a storage room provided by the local government , in coordination with the research group . Community meetings were held weekly in each village , with groups of 10–20 households , to provide Chagas disease awareness ( including instructions for entomological monitoring ) , at the end of which households were offered to enroll in the intervention . Each attending family was also asked to bring another household to be enrolled at the subsequent meeting ( neighbor or family member ) to achieve a snowball effect and increase enrolment in the community . Thus , any construction in the community that was used as a house and sleeping quarter was eligible to participate in the intervention , and only constructions used for other purposes were considered ineligible ( stores , offices , school , or abandoned houses ) . Households were then visited by the carpenters for measurements of window size and type . The social workers and the carpenters coordinated with the research team to ensure the provision of materials for screen manufacture , which were stored and administered under their supervision [26] . Screens were then installed by the carpenters in coordination with the households . Once implementation was initiated , the research team played a minimal role in these activities , which leadership was effectively transferred to the social workers [26] . However , weekly supervision of installed screens was performed to evaluate coverage . Maps of participating households were elaborated in QGis [27] to assess their geographic distribution . Because house location in the periphery and the proximity of public lights have been found to be risk factors for infestation [22 , 23 , 28] , we measured the distance of houses participating in the intervention and of control house to the periphery of the village and to the nearest public light to assess any bias between these two groups . Following implementation of the intervention , its effects on house infestation by triatomines was monitored for three consecutive seasonal infestation periods to assess it efficacy and sustainability . A subset of houses was randomly selected for detailed follow-up , 95 to 122 with insect screens and 77 to 147 without insect screens , depending on the year and distributed evenly in the villages ( S1 Fig ) . Control houses were selected from the same villages where screens had been installed , as well as from a third village ( Bokoba ) where no vector-control intervention had been implemented . Entomological monitoring was performed by community participation , as reported before [18] , during 2 weeks for each village and each year during the infestation season . Briefly , households were instructed to collect any bugs they would detect inside or outside their own house while performing their normal daily activities . They were trained for the safe handling of triatomines to avoid infection , but no additional training was provided since inhabitants are rather knowledgeable of the bugs [24] . This method is more sensitive than timed manual searches by trained research personnel for the detection of low level infestation [18] . Nonetheless , because of potential difference in household participation between houses with screen and houses without screens , as well as changes in motivation over time , we also used mouse-baited traps [29] for unbiased monitoring of domestic infestation . A total of eight traps per house were used , consisting of four traps per bedroom window , with two traps on the inside and two traps on the outside of each screen/window ( S2 Fig ) . Often , however , one of the outside traps and one of the inside traps were set on the outside and on the inside of the door , respectively . Traps were used only one night per house . They were set up in the evening and removed in the morning after collecting the bugs caught on the traps . We evaluated house infestation index ( inside and outside infestation ) expressed as the percentage of houses with bugs , and bug density index ( inside and outside bug abundance/house ) expressed as number of bugs/house , as well as the proportion of inside infestation and inside bug abundance/house . We adopted a ‘Force of Infection’ ( ‘FoI’ ) approach ( see [30] for a review of ‘FoI’ models of T . cruzi transmission ) to predict the impact of the reduction in vector abundance on the risk of transmission in the studied villages . Such modelling allows calculating the prevalence and the incidence in human with respect to the so-called FoI , i . e . the per unit time probability for a susceptible human to become infected , that is commonly modelled as follow; λ=1− ( 1−p ) C where C stands for the number of potentially infectious contact ( PIC ) between a susceptible individual and infected vectors , and where p denotes the probability of transmission per PIC . The former can in turn be calculated using the typical relationship depicting vector ecology; C=Nv*pv*b*fhNh where Nv and Nh denote the abundance of vectors and humans , pv stands for the prevalence of T . cruzi infection in vectors , while b and fh are the vector biting rate and the proportion of those bites that are made on humans . Under the assumption that both the force of infection and the human death rate , μ , are constant , the prevalence in human is given by: ph=λλ+μ while the incidence was calculated as the product of the force of infection ( λ ) and the number of inhabitants . We parameterized those equations using the same estimates as in Nouvellet et al . [31] but for the abundance of bugs that we estimated to be modified by the same proportion as that of inside infestation in houses with screens , in order to calculate the changes in incidence and prevalence T . cruzi infection due to the intervention . We also calculated the cost of the intervention , which included materials and labor for screen manufacture and installation , but not the labor from other stakeholders such as social workers or local governments . A random subset of insect screens was also inspected every year to evaluate their integrity/wear over time ( N = 182–244 depending on the year ) . Visual inspection allowed to classify each screen as ( i ) in perfect condition , ( ii ) with minor damage/wear ( some small holes or tears but still effective ) , and ( iii ) major damage ( major holes or tears , screen partially or totally removed , major loss of efficacy ) . Entomological data are expressed as infestation index and density index , or proportion , and are presented as mean ± 95% confidence intervals . Statistical comparisons between houses with insect screens and control houses were made by χ2 tests for the infestation index and by Z-tests ( comparing the mean between two Poisson distributions ) for the density index . Distance data are presented as mean ± SEM and are compared with t tests . An informed consent was signed by the head of each participating household prior to inclusion in the study and performing any of the activities described . The project was approved by the institutional bioethics committee of the Autonomous University of Yucatan and of the TDR/World Health Organization .
A total of 1 , 606 window screens were installed in 822 households in the villages of Teya and Sudzal , over a period of about 10 months ( Fig 1A ) . Logistical difficulties to ensure the delivery of materials to the carpenters at the beginning of the intervention caused some minor delays , as well as some difficulties in coordinating the carpenter workshop with the households for screen installation at the end of the intervention . Nonetheless , all workshops generally produced about 50–100 screens/month ( Fig 1A ) . This included an initial visit to each participating household for measurement of window size and type , manufacture and installation of the screens . Maximum manufacture and installation rates even reached up to 150 screens/months in two occasions in Sudzal . This rate of implementation seems very good given the rudimentary facilities and limited personnel ( family members ) available in each workshop ( Fig 1B ) and the various styles and designs of the screens so that they appropriately fitted each window ( Fig 1C ) . The 822 households participating in the study corresponded to a coverage of the intervention of 59% of households in Teya ( 417/702 ) and 58% in Sudzal main village ( 295/509 ) . However , because some households already had insect screens , and other were ineligible for the study because they were not used for sleeping ( stores , offices , or abandoned ) , a higher proportion of houses in each village actually had screens . For example this proportion reached 76% of houses in Sudzal main village ( 336/440 eligible houses ) . Participating households were distributed across the entire village in both Teya and Sudzal ( Fig 2 ) , and thus were exposed to a comparable risk of infestation as non-participating households without insect screens , due to their location within the village [23 , 28] . Indeed , houses with screen intervention were located at a distance of 119 ± 10 m of the periphery of the village , while control houses not participating in the intervention were located at 114 ± 9 m of the periphery ( t test , P = 0 . 52 ) . Similarly , houses from the intervention were located at a distance of 24 . 6 ± 2 . 1 m from public lights , while control houses not participating in the intervention were located at 23 . 7 ± 3 . 6 m from public lights ( t test , P = 0 . 58 ) . Entomological surveillance was used to evaluate the efficacy of the intervention to reduce triatomine presence for three years following the installation of screens . Bugs were collected by community participation during 2 weeks per village and year of follow-up in both houses with screens and control houses that did not participate in the intervention . Houses from the village of Bokoba , which did not receive control intervention , were used as additional controls . Furthermore , mouse-baited traps were also used to complement community bug surveillance and detect potential bias in community participation . A total of 5744 traps were set inside and outside 718 houses ( one trapping night/house ) evaluated during the three-year follow-up ( see Materials and Methods ) . A total of 440 bugs were collected over the three years of follow-up . Most bugs were collected by community participation ( 368 bugs ) compared to mouse-baited traps ( 72 bugs ) . Ninety six percent were adults ( 63% females and 37% males ) , and only 4% were nymphal instars as observed before [17 , 19] . A total of 147 bugs were collected in houses without screens ( both outside , i . e . front or backyard , or outside walls of the houses and inside the houses ) , while 220 were collected in houses with insect screens ( both outside and inside ) . The overall infestation index ( i . e . taking into account bugs collected inside and outside houses ) was 18 . 6% in 2013 ( 34/183 ) , 22 . 5% in 2014 ( 59/262 ) , and 30 . 4% in 2015 ( 83/273 ) , which likely reflected our multiple efforts to motivate and increase community participation to find bugs during the study . There were no significant differences in infestation among control houses from the villages of Teya , Sudzal and Bokoba ( P = 0 . 538 ) , which data were thus pooled for every year of follow-up . Based on bug collections by community participation only , overall infestation index ( inside + outside ) indicated that around 15 . 3–26 . 4% of houses were infested over the three years study period , with no significant differences in infestation between houses with and without screens , except in year 3 which showed a significant increase in infestation in houses with insect screens ( Fig 3A ) . Triatomine density index ( inside + outside bugs ) showed a small but significant increase from 0 . 34 to 0 . 56 bug/house for houses without screen and houses with insect screens , respectively , for the 3 years of follow-up ( Fig 3B ) . This increase was due to a large increase in the density of bugs collected from houses with screens in year 3 of the intervention . On the other hand , data from mouse-baited traps evidenced no significant difference in infestation index ( inside + outside bugs , Fig 3C ) and density index ( inside + outside , Fig 3D ) between houses without screens and houses with insect screens for any of the years of follow-up . As this method provides an unbiased evaluation of infestation , the increase in infestation observed by community participation in year 3 is very likely to be due to a much greater awareness of Chagas disease and motivation to participate of households who benefited from the insect screens . Indeed , we did not expect the presence of screens to reduce overall triatomine populations since these screens were not impregnated with insecticide and thus ineffective at killing bugs . We then evaluated in more details if infestation occurred inside or outside houses , to evaluate if insect screens helped keeping bugs outside of houses . Based on community collections , both inside infestation and density indexes were constant over time with negligible yearly variations in control houses ( S3A and S3B Fig ) . On the other hand , in houses with screens , both inside infestation and density index were lower than in control houses for years 1 and 2 , but higher in year 3 . Based on traps ( S3C and S3D Fig ) , the effect of screens on inside infestation and density indexes could not be detected in years 1 and 2 , but both indexes tended to be lower in houses with screens in year 3 , strongly suggesting that the increase shown by community participation for that year is most likely due to and increased community participation of this group due to greater Chagas disease awareness . Therefore , we also assessed the efficacy of the intervention by looking at the proportion of bugs inside/outside , to control for potential bias in surveillance effort among households . Based on both community participation and mouse-baited trap bug collections , there was a small but statistically significant decrease in the proportion of houses with indoor infestation when screens were present , and a constant trend was observed each year of the follow-up period ( Fig 4A ) . More importantly , the proportion of bugs inside houses was reduced in houses with insect screens , and this effect was statistically significant for two of the three years of follow-up ( Fig 4B ) , as well as for the cumulative three years . Thus , the presence of insect screens effectively kept more bugs on the outside of houses . To further assess the impact of this intervention on Chagas disease in these communities , we used a previously developed model linking entomological data to the prevalence of Trypanosoma cruzi infection in humans [31] . With this model , the interventions proposed in this study predicted to lead to a 32% reduction in yearly incidence and in the prevalence of T . cruzi infection . Based on a current seroprevalance of 1–4% in these villages [7] , we could thus expect to reduce it to 0 . 7–2 . 7% with this vector control intervention . In terms of costs , the average cost per insect screen was of 17 $US , i . e . an average cost of 34 $US per household for the coverage of one bedroom with 2 windows . About half of this cost corresponded to labor , and the other half for materials , most of which being for the wood needed for the frames of the screens ( 70% of material costs ) . We also evaluated the integrity of the screens over time , as a preliminary assessment of the long-term sustainability of the intervention . After three infestation seasons , only about 10% of the screens presented major damage to render them ineffective , and the large majority remained in perfect conditions ( Fig 5 ) .
Vector control against Chagas disease remains key to prevent new cases of T . cruzi infection , but has proven challenging for triatomine species behaving as intrusive bugs which transiently invade houses [9] . Previous modeling and pilot field work showed that insect screens acting as a physical barrier preventing bugs from entering houses could be a valuable control strategy in this situation [11 , 12] , which raised questions for the scaling-up and implementation of this kind of intervention . We investigated here the implementation and efficacy of a village-scale intervention involving the manufacture and installation of insect screens covering two rural communities in the Yucatan , Mexico , under a community participation framework . We were able to reach a high participation and coverage in the two study villages , reaching over 70% of eligible houses within 10 months . Such a level of coverage can be considered excellent for a community-based intervention as it reached comparable levels as a systematic vertical intervention [32–34] . Engaging the community and all stakeholders in Chagas disease awareness was likely a fundamental activity allowing to reach such a high participation and coverage of the intervention . The high acceptance and desirability of insect screens as perceived by households also likely contributed [24] . In term of efficacy of the intervention to control triatomine infestation , we observed that it had no significant effect on overall ( inside and outside ) house infestation or bug density index , as can be expected from an intervention based on a physical barrier , which does not kill any bugs . There was even an increase in infestation and bug density in the third year of follow-up in houses with screens based on bug collections by community participation . The most likely explanation for his observation is that there was a much greater interest and thus more active entomological surveillance from households with screens compared to households who did not participate/accept the screens . This explanation is further supported by bug collection data with mouse-baited traps , which provide an unbiased assessment of triatomine infestation , and confirmed that there were no differences of the total density of bugs found inside and outside between houses with and without insect screens . However , when we analyzed infestation in further detail and assessed if bugs were found inside or outside houses , it became clear that insect screens effectively worked as physical barriers limiting bug entry inside houses and maintaining them outside . Importantly , with only two windows covered by screens , we observed a significant reduction in bug abundance inside houses , and we estimated that the concomitant reduction in vector-host contacts would translate into a 32% reduction in yearly incidence and in the prevalence of T . cruzi infection in humans . Such a reduction is highly significant considering that doors ( which often remain open at this very hot time of the year ) , and in some cases additional windows , were left unprotected . We can thus expect that a more complete coverage of houses with screens on all openings would allow reaching an even greater efficacy . Indeed , our previous pilot testing of insect screen covering all windows allowed tor reach 87–100% efficacy in reduction in triatomine collections over a 2 years follow-up [12] . In terms of cost , we reached an average of 17 US$ per screen , although this varied according to size and type of screen , corresponding to about 34 US$ per household . This is somewhat lower than the cost of screens used in our previous pilot study [12] , and can be attributed to the bulk prices for materials that we were able to obtain for large-scale purchases . Since most houses have a total of 3–4 windows when considering all rooms [22] , the cost for a complete coverage of all the windows would reach 51–68 US$ . Importantly , this is of a comparable magnitude to what families currently spend on various domestic insecticide such as plug-in repellent , repellent coils , or insecticide sprays , which amount to an average of 32 $US/year [24] . This suggest that while one of the main barrier for families to purchase their own insect screen is their perceived excessive cost , specific education may help them reconsider and redirect their spending from insecticide products to insect screens [24] . The long durability of the screens observed in our study further indicate that the proposed intervention would be highly sustainable , allowing to maintain a barrier effect for many years with minimal screen maintenance . Indeed , while no specific instructions were provided to households for the care and maintenance of their screens , the large majority remained in perfect condition after three years . Screens would likely need to be replaced only every 5–10 years , making the intervention very cost-effective . In addition such insect screens would not only allow for the control of triatomine bugs , but would also be effective against mosquitoes , thus preventing Dengue , Chickungunya , or Zika virus infections [35] , which are also major public health issues in the region . Therefore , this intervention could be part of an integrated program for vector-borne disease control , further increasing its cost-effectiveness . Our study presents nonetheless some limitations . First the intervention allocation was not randomized , as our objective was to cover the entire community in each village , as this provided some estimates of potential coverage of the community-based enrollment for further scaling up of the proposed intervention . Also , only bedrooms were considered for the installation of screens , due to limited resources , which may have lead to some underestimation of screen efficacy . Finally , the increased bug collections in intervened houses that we interpreted as an increase in awareness of these households generated some bias for the evaluation of entomological efficacy of the insect screens , that needed to be controlled for . On the other hand , this also provided evidence of the successful mobilization of the community for more active vector surveillance . In conclusion , the Ecohealth approach that we developed here and the proposed community intervention based on the installation of insect screens appears to be effective for the long-term and sustainable control of invasive T . dimidiata vectors in the Yucatan peninsula , Mexico . This approach and strategy may also be easily adapted to other invasive triatomine species as well as other regions/countries with comparable eco-epidemiological settings , and would be an excellent component of a larger integrated program for the control of a variety of other vector-borne diseases , bringing additional benefits to the communities . Therefore , our results should encourage a further scaling-up of our implementation strategy in additional communities in the region . The additional targeting of additional risk factors for triatomine infestation , such as public lights or the presence of domestic animals may be explored to further increase the efficacy of the intervention .
|
Intrusive triatomine bugs such as Triatoma dimidiata in the Yucatan peninsula , Mexico , are responsible for the transmission of Trypanosoma cruzi to humans , which can lead to Chagas disease . The control of these bugs is a challenge as insecticide spraying is poorly effective , and alternative strategies need to be developed for a sustainable control . We tested a novel Ecohealth approach , based on window insect screens manufacture and installation through community participation to reduce the presence of bugs inside houses . The proportion of triatomines found inside houses was significantly reduced in houses with insect screens , which effectively kept more bugs on the outside of houses . We estimated that the intervention would lead to a 32% reduction in new cases of infection each year and in the prevalence of T . cruzi infection . The low cost of the intervention and the durability of screens further indicate that the proposed Ecohealth approach is effective for the long-term and sustainable control of intrusive T . dimidiata vectors in the Yucatan peninsula , Mexico . This strategy may also be easily adapted to other bug species and regions with similar characteristics and sould be expanded in additional villages in the region .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"tropical",
"diseases",
"vector-borne",
"diseases",
"parasitic",
"diseases",
"animals",
"parasitic",
"protozoans",
"protozoans",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"insect",
"vectors",
"zoology",
"infectious",
"diseases",
"agrochemicals",
"protozoan",
"infections",
"disease",
"vectors",
"insects",
"agriculture",
"arthropoda",
"insecticides",
"trypanosoma",
"cruzi",
"trypanosoma",
"eukaryota",
"chagas",
"disease",
"entomology",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"organisms"
] |
2018
|
Non-randomized controlled trial of the long-term efficacy of an Ecohealth intervention against Chagas disease in Yucatan, Mexico
|
Professional phagocytes generate a myriad of antimicrobial molecules to kill invading microorganisms , of which nitrogen oxides are integral in controlling the obligate intracellular pathogen Leishmania . Although reactive nitrogen species produced by the inducible nitric oxide synthase ( iNOS ) can promote the clearance of intracellular parasites , some Leishmania species/stages are relatively resistant to iNOS-mediated antimicrobial activity . The underlying mechanism for this resistance remains largely uncharacterized . Here , we show that the amastigote form of L . amazonensis is hyper-resistant to the antimicrobial actions of cytokine-activated murine and human macrophages as compared to its promastigote counterpart . Amastigotes exhibit a marked ability to directly counter the cytotoxicity of peroxynitrite ( ONOO− ) , a leishmanicidal oxidant that is generated during infection through the combined enzymatic activities of NADPH oxidase and iNOS . The enhanced antinitrosative defense of amastigotes correlates with the increased expression of a tryparedoxin peroxidase ( TXNPx ) isoform that is also upregulated in response to iNOS enzymatic activity within infected macrophages . Accordingly , ectopic over-expression of the TXNPx isoform by L . amazonensis promastigotes significantly enhances parasite resistance against ONOO− cytotoxicity . Moreover , TXNPx-overexpressing parasites exhibit greater intra-macrophage survival , and increased parasite growth and lesion development in a murine model of leishmaniasis . Our investigations indicate that TXNPx isoforms contribute to Leishmania's ability to adapt to and antagonize the hostile microenvironment of cytokine-activated macrophages , and provide a mechanistic explanation for persistent infection in experimental and human leishmaniasis .
Leishmania spp . are the causative agent of the neglected tropical disease leishmaniasis , causing significant morbidity and mortality worldwide . Leishmaniasis presents with a broad spectrum of clinical symptoms , ranging from self-healing cutaneous lesions to life-threatening systemic disease that is dependent on a complex interaction between the infecting parasite species and the host immune responses . Leishmania exhibit a digenetic lifecycle , converting from a flagellated promastigote stage in the sand fly vector to a non-motile amastigote stage within vertebrate host macrophages ( MΦs ) . Amastigotes that are responsible for the disease survive and replicate in the hostile intracellular environment of MΦs . Oxygen-dependent antimicrobial defenses , encompassing both reactive oxygen and nitrogen species ( ROS and RNS , respectively ) , are extensively studied innate host defense mechanisms used against invading microorganisms . Upon phagocytosis of pathogens , MΦs generate superoxide ( O2•− ) via the NADPH phagocyte oxidase respiratory complex . O2•− is subsequently dismutated enzymatically or spontaneously to hydrogen peroxide ( H2O2 ) , an oxidant that readily diffuses across membranes and has the propensity to generate the highly reactive hydroxyl radical ( OH•− ) through its reaction with iron . [1] . In addition , the diatomic radical NO , which is generated by iNOS during innate recognition of pathogen-associated molecular patterns or in the IFN-γ-primed response , is an intrinsic component of host defense to infection [2] . It is likely that the anti-Leishmania activity emanating from iNOS involves a myriad of NO congeners , including nitrogen dioxide ( NO2 ) , dinitrogen dioxide ( N2O3 ) , and peroxynitrite ( ONOO− ) , which are generated in the reaction of NO with O2 and O2•− . In particular , the strong oxidant ONOO− has been shown to mediate Leishmania toxicity [3] , most likely through the nitration/and or oxidation of parasite membrane proteins [4] . These molecules are crucial for controlling the parasite burden during the course of cutaneous and visceral leishmaniasis [5]–[7] . It is also known that both ROS and RNS can mediate cytotoxicity through the oxidation or S-nitrosylation of redox-active cysteine thiols in proteins essential for cell function [8] , [9] , and that the leishmanicidal effect of NO can partially be attributed to inactivation of the cysteine proteinase virulence factor through S-nitrosylation of the Cys25 catalytic residue [10] , [11] . In order to survive in the harsh phagolysosomal compartment of MΦs , Leishmania have developed a distinct system for defense against host-derived reactive species . These parasites are known to actively decrease production of ROS/RNS by interfering with NADPH oxidase assembly [12]–[14] , or through the disruption of signaling pathways governing iNOS transcription [15] . They also can enzymatically detoxify reactive species generated endogenously during parasite metabolism , or exogenously via the anti-parasite immune response in the sandfly midgut or within professional phagocytes . Leishmania possess several antioxidant systems , including superoxide dismutases , peroxidases , and low-molecular weight thiols [16] . In contrast to higher eukaryotes , Leishmania do not express catalases or selenium-containing glutathione peroxidases; instead , they rely on the dithiol trypanothione to maintain their redox homeostasis . Trypanothione transfers reducing equivalents to the downstream antioxidant enzyme couple tryparedoxin ( TXN ) /tryparedoxin peroxidase ( TXNPx ) [17] , forming a key antioxidant cascade vital for Leishmania resistance to ROS/RNS [18] , [19] . The TXNPx enzymes are two-cysteine peroxiredoxins that act as homodimers to detoxify H2O2 , organic hydroperoxides , ONOO− , and NO , and are highly conserved among the kinetoplastids [20]–[22] . Leishmania encode at least two TXNPx isoforms: one cytosolic form ( a 199-amino acid protein with a predicted molecular mass of 20 . 1 kDa ) and a mitochondrial form ( a 226-amino acid protein with a predicted mature molecular mass of 21 . 4 kDa ) [23] . These enzymes can promote Leishmania virulence due to their antioxidant and chaperone activities , respectively [24] , [25] . In addition , a third developmentally-regulated isoform that is preferentially expressed by the mammalian parasite stage , termed TXNPx1 or Pxn1 , has been identified [21] , [26] . Collectively , these trypanothione-dependent processes are essential for Leishmania survival [27] , [28] , and are implicated in the development of parasite drug resistance [29] . Therefore , these pathways are actively being investigated as potential therapeutic targets for treatment of leishmaniasis . Interestingly , susceptibility to ROS and RNS is highly variable among Leishmania species and developmental stages , even though genomic data indicate that all sequenced species possess similar antioxidant systems . For example , members of the L . mexicana complex ( L . mexicana , L . amazonensis , L . pifanoi ) exhibit enhanced resistance to direct treatment with oxyradicals and nitrogen oxides in vitro [30] and increased intracellular survival in cytokine-activated MΦs [31] . Parasites of the L . mexicana complex often cause persistent cutaneous lesions in humans and induce non-healing cutaneous lesions in BALB/c , C3H , and C57BL/6 strains of mice , suggesting that enhanced intra-MΦ survival is an important factor in host susceptibility . However , the mechanisms underlying the increased resistance of certain Leishmania species or stages to the antimicrobial actions of both oxyradicals and nitrogen oxides are unknown . In this study , we compared the ability of L . amazonensis promastigotes and amastigotes to resist the antimicrobial activity of MΦs that were pre-activated with cytokines to boost their killing capacity . In addition , we evaluated the susceptibility of both parasite stages to authentic ROS and RNS , which are likely to be encountered by Leishmania in the parasitophorous vacuole . Our investigations indicate that while promastigotes are readily killed by cytokine-activated MΦs , axenic amastigotes are not only resistant to MΦ-mediate killing , but also grow intracellularly in spite of elevated host NO synthesis . In addition , amastigotes are more resistant to the leishmanicidal activity of RNS compared to promastigotes . Resistance of amastigotes to MΦ-mediated killing correlates with the increased expression of a TXNPx isoform by this stage of the parasite , ( a 190-amino-acid protein with a predicted molecular mass of 20 kDa similar to L . chagasi Pxn1 [22] . Our findings uncover a novel Leishmania antioxidant/antinitrosative defense strategy that antagonizes iNOS enzymatic activity and fosters intracellular parasite fitness and virulence mechanisms . They also help explain why certain Leishmania species of the L . mexicana complex , including L . amazonensis , cause persistent disease in spite of Th1-like immune responses in infected or vaccinated hosts .
Female wild-type C57BL/6 ( B6 ) , congenic iNOS-deficient ( iNOS−/− ) mice , and BALB/c mice were purchased from Taconic Laboratories ( Hudson , NY ) . Mice were maintained under specific pathogen-free conditions and used at 6–10 weeks of age following protocols that have been approved by the Institutional Animal Care and Use Committee ( protocol #9803016 ) at the University of Texas Medical Branch ( UTMB ) in Galveston , TX . UTMB complies with the USDA Animal Welfare Act ( Public Law 89-544 ) , the Health Research Extension Act of 1985 ( Public Law 99-158 ) , the Public Health Service Policy on Humane Care and Use of Laboratory Animals , and the NAS Guide for the Care and Use of Laboratory Animals ( ISBN-13 ) . UTMB is a registered Research Facility under the Animal Welfare Act , and has a current assurance on file with the Office of Laboratory Animal Welfare , in compliance with NIH Policy . C57BL/6 mice were inoculated s . c . with 5×106 stationary-phase promastigotes in the right hind foot . Lesion development was evaluated biweekly for 9 weeks , at which time the mice were sacrificed for determination of parasite burden by using qRT-PCR as described below . Infectivity of L . amazonensis ( RAT/BA/74/LV78 ) was maintained by regular passage through BALB/c mice . Promastigotes were cultured at 26°C in Schneider's medium ( Invitrogen , Carlsbad , CA ) , pH 7 . 0 , supplemented with 20% fetal bovine serum ( FBS ) ( Hyclone , Logan , UT ) , and 50 µg/mL gentamicin ( complete Schneider's medium ) . Axenic amastigotes were maintained at 32°C in Grace's medium ( Invitrogen ) , pH 5 . 3 supplemented with 20% FBS . Promastigote growth in complete Schneider's medium was evaluated daily by direct counting of parasites using a hemacytometer . Metacyclic promastigote forms were purified using the 3A . 1 monoclonal antibody , as previously described [32] . Tissue-derived amastigotes were harvested from foot tissues of infected BALB/c mice ( ∼12 wk post-infection ) and cultured at 32°C for 48 h before use . To generate the TXNPx1-overexpressing strain , the TXNPx1 open-reading frame was amplified from L . amazonensis genomic DNA by using the forward 5′-AAAACCCGGGACCATGTCCTGCGGTGACGCCAA-3′ and reverse 5′-AAAACCCGGGTCACTTATTATGGTCGACCTTCAGGCCAGG-3′ primer set ( SmaI restriction sites underlined ) , and directionally cloned into the SmaI site of pXG to generate pXG::TXNPx1 . The orientation of the TXNPx1 open-reading frame was confirmed by PCR and sequencing . pXG and pXG::TXNPx1 were electroporated into logarithmic phase promastigotes by using the high-voltage ( 1500V ) method , as previously described [33] . Stably transfected parasites were selected by growth in Schneider's medium containing G1418 ( 50 µg/mL ) . Confirmation of TXNPx1 overexpression was determined by immunoblotting . Promastigote and amastigote cultures carrying the episomal pSP72–YNEO-αIR-Luc1 . 2 [34] or pXG vectors were supplemented with G1418 . Axenic amastigote or stationary promastigote cultures under ten passages were routinely used for in vitro infections . Bone marrow-derived MΦs were generated from C57BL/6 and congenic iNOS−/− mice , as previously described [35] , and used for experimentation after 9 days of incubation with M-CSF ( 20 ng/mL ) . MΦs were seeded in 24-well culture plates ( 3×105 cells/well ) and allowed to attach overnight . The THP-1 monocytic cell line was activated with PMA ( 50 ng/mL ) for 3 days prior to infection . Where indicated , MΦs and THP-1 cells were activated with LPS ( 100 ng/mL ) and IFN-γ ( 100 U/mL ) for 16 h prior to infection , which were maintained in the medium for the duration of the experiment; these cells are referred to as activated MΦs in all our experiments herein . The anti-leishmanial activity of MΦs and THP-1 cells was evaluated after challenge with a multiplicity of infection of 2 stationary-phase promastigotes or axenic amastigotes . After 1 h of infection , cells were washed with warm PBS to remove extracellular parasites , and new medium was added . Total RNA was harvested from infected MΦs , and cDNA was used in qPCR reactions with parasite-specific primers ( forward 5′-AACGTGAACAACTGGATGTGCGTC-3′ and reverse 5′-ATGGTACCAAGCTTGACACATGCC-3′ ) directed against the single copy ubiquitin hydrolase ( Ubiq ) gene known to be constitutively expressed in both developmental stages [32] . Alternatively , genomic DNA was harvested from infected MΦs and used as a template in qPCR reactions with the L . amazonensis cysteine proteinase ( LaCys ) -specific primer set ( forward 5′-TCGTGCTGGGCCTTCTC-3′ and reverse 5′-TTGCAGCCCACTGACCTT-3′ ) . The relative parasite load was calculated by using the ΔΔCt method normalizing parasite Ubiq or Cys to MΦ Gapdh levels amplified with the murine specific forward 5′-GAGCTGAACGGGAAGCTCAC-3′ and reverse 5′-ACCACCCTGTTGCTGTAGC-3′ primers . Intracellular survival is expressed as the parasite burden determined at the indicated time relative to the parasite burden after 1 h of internalization . Nitrite in the MΦ culture supernatant was measured by the Griess reaction , and concentrations determined by regression analysis were compared to known nitrite standards . For secondary infections , amastigotes were released from infected cells by using 0 . 01% SDS in PBS , as previously described [36] . MΦ-derived amastigotes were counted , and equal parasite numbers ( MOI 2 ) were then used to infect secondary activated MΦs . Parasite load was determined 48 h post-infection by qPCR . Total RNA was isolated from lesions or 1×106 infected MΦs by using the Qiagen RNeasy kit ( Qiagen , Valencia , CA ) , and cDNA was synthesized by using 1 µg total RNA and iScript reverse transcriptase ( Biorad , Hercules , CA ) . Real-time PCR was performed by using a BioRad iCyler iQ Real-Time PCR System ( BioRad ) . The TXNPx1 cDNA was amplified by using the forward 5′-ACCGCGGTCTCTTCATCATCGACCC-3′ and reverse 5′-TCACTTATTATGGTCGACCTTCAGGCCAGG-3′ primers . The cycle threshold ( Ct ) value for parasite TXNPx1 or Ubiq mRNA was determined . Total parasite load in lesions was calculated using regression analysis based on Ubiq mRNA levels compared to a standard curve generated with amastigote-derived cDNA . Calculations were made considering 1 pg cDNA was equal to 7 parasites . Lesion TXNPx1 mRNA levels were normalized to Ubiq mRNA levels to correct for parasite load , and TXNPx1 expression levels in pXG control promastigote-infected lesions were set to 1 . The leishmanicidal activities of the NO donor spermine NONOate ( sNO , Cayman Chemical , Ann Arbor , MI ) , polyamine spermine ( a negative control for sNO ) , hydrogen peroxide ( H2O2 ) , or peroxynitrite ( ONOO− ) were monitored by assessing luciferase activity in parasites that carry the episomal vector pSP72–YNEO-αIR-Luc1 . 2 [34] . sNO was assumed to exhibit a half-life of ∼100 minutes under our experimental conditions , based on its temperature- and pH-dependent NO release [37] . ONOO− was freshly synthesized and quantified using UV-absorbance spectroscopy ( ε302 nm = 1670 M−1 cm−1 ) , as previously described [38] . Since the parasitophorous vacuole is thought to be a nutrient limiting environment , we opted to perform the cytotoxicity assays in PBS in the absence of any exogenous metabolites . Neither parasite stage showed any loss of viability after 4 h of incubation in PBS alone ( not shown ) . Bulk stationary-phase promastigotes or axenic amastigotes were diluted in PBS alone to a final concentration ( 1×106 parasites/mL ) . For comparing susceptibility of promastigotes and amastigotes , 1×105 parasites were aliquoted into 96-well plates and treated with the indicated concentrations of sNO , H2O2 , or ONOO− at 32°C for 4 h . Samples were lysed with luciferase assay buffer , and luciferase activity was determined using a luciferase assay system ( Promega , Madison , WI ) and measured as relative light units ( RLU ) on a Veritas microplate luminometer . The results are expressed as luciferase activity in treated samples relative to luciferase activity in untreated controls ×100% . For comparing susceptibility of transfected promastigotes , stably transfected stationary-phase promastigotes that carry either pXG or pXG::TXNPx1 were diluted in PBS ( 1×106 parasites/mL ) , and 1×105 parasites were seeded into 24-well plates and treated with the indicated concentrations of ONOO− at 26°C for 4 h . After the treatment , 0 . 8 mL of complete Schneider's medium was added , and surviving parasites were cultured for 4 days . Parasites were enumerated by counting under a hemacytometer . Results are presented as parasites/mL in treated relative to untreated samples ×100% . Whole cell lysates were prepared from 1×108 stationary-phase promastigotes , lesion-derived or axenic amastigotes , or 1×106 infected MΦs . Samples normalized to 2 µg ( parasite only lysate ) or 10 µg ( infected MΦ lysate ) total protein were resolved by using 12% ( v/v ) SDS-PAGE , electrophoretically transferred to a nitrocellulose membrane , and immunoblotted with anti-TXNPx1 ( generated by inoculating mice with the purified recombinant C-terminal 40 amino acids of the L . pifanoi truncated TXNPx1 isoform ) , anti-β actin ( Sigma , A5441 ) , anti-α tubulin ( Sigma , T9026 ) , and/or anti-luciferase ( Sigma , L2164 ) . α-tubulin was used as a loading control for comparing protein expression between promastigote samples . However , α-tubulin expression was relatively low in amastigotes ( data not shown ) , presumably due to their lack of flagella , and anti-β actin was not a good loading control for lesion-derived amastigotes ( data not shown ) . Therefore , we opted to use Ponsceau S staining as a loading control when comparing promastigote and amastigote protein expression levels . Band intensity was measured by using ImageJ software [39] . One-way ANOVA was used for multiple group comparisons . Data from time course and titration experiments were evaluated by using a two-way ANOVA followed by a Bonferroni post-test . Differences between individual treatment groups were determined using a Student's t-test . A p value≤0 . 05 was considered statistically significant ( GraphPad Software , San Diego , CA ) .
MΦs readily phagocytose both promastigotes and amastigotes . However , our understanding of how the insect- or mammalian-stage of the parasite subverts the innate immune response of MΦs is incomplete . Most in vitro studies have used microscopic counting of parasites and host cells or conventional PCR assays to determine parasite burden , which have intrinsic limitations such as low sensitivity and high variation . For a more reproducible examination of Leishmania parasite survival and intracellular burden in control and cytokine-activated MΦs , we used qRT-PCR to quantify the amount of parasite RNA relative to host RNA in infected samples . In addition , luciferase-expressing promastigotes and amastigotes were used for some infections ( Figure S1 ) , allowing luciferase activity in infected MΦ lysates to be quantified and correlated to intracellular parasite loads . These molecular-based methods eliminate potential bias that exists with traditional microscopic counting methods commonly used to determine intracellular Leishmania load in in vitro infection studies . Upon infection of resting MΦs , both L . amazonensis promastigotes and amastigotes replicated efficiently with parasite loads increasing approximately 10-fold after 4 days of infection ( Figures 1A and B , open symbols ) . However , we consistently found that ∼95% of ingested promastigotes were killed in IFN-γ/LPS-activated MΦs within 24 h of infection ( Figure 1A and Figure S1A , closed symbols ) . The surviving parasite population began replicating in cytokine-activated MΦs at 48 h post-infection and continued to expand for the duration of the experiment ( Figure 1A ) . In sharp contrast , amastigotes were resistant to MΦ microbicidal mechanisms ( Figure 1B and Figure S1B , closed symbols ) . Amastigote replication in cytokine-activated MΦs was limited during the first 24 h of infection , but growth proceeded with similar kinetics in both control and activated cells thereafter ( Figure 1B ) . Leishmania are susceptible to the antimicrobial actions of RNS produced by the host iNOS hemoprotein [7] , [40] . Therefore , we evaluated the ability of bone marrow-derived MΦs from C57BL/6 and iNOS−/− mice to kill L . amazonensis promastigotes . In contrast to activated wild-type MΦs that efficiently killed ∼95% of parasites , primed iNOS−/− MΦs only eliminated ∼50% of engulfed promastigotes ( Figure 1C ) . This suggested that MΦ killing of promastigotes was largely dependent on iNOS; however , other MΦ effector mechanisms independent of NO clearly contribute to promastigote killing in the initial stages of infection when host cells are pre-activated [5] . Reportedly , Leishmania parasites can interfere with iNOS activity , thereby decreasing the effective NO concentration in infected cells [15] . To determine if the increased amastigote resistance to MΦ-mediated killing was due to a defect in RNS production , we measured the nitrite concentration in culture supernatants of infected cytokine-activated MΦs . Both promastigote- and amastigote-infected MΦs produced comparable levels of RNS at the time of infection ( Figure S2A ) and continued to produce similar concentrations of RNS up to 48 h post-infection ( Figure S2B ) , suggesting that neither L . amazonensis promastigotes nor amastigotes interfere with iNOS production or activity in these primed cells . Given that promastigote clearance in cytokine-activated murine MΦs was largely dependent on iNOS activity , and that human MΦs tend to produce less NO compared to murine cells [41] , we evaluated the ability of primed human THP-1 monocytes to eliminate promastigotes and amastigotes . Similar to murine MΦs , luciferase-expressing amastigotes were resistant to THP-1 antimicrobial activity , while a significant percentage of internalized promastigotes were killed after 2 d of infection ( Figure S1C , p<0 . 01 ) . Together , these data indicate that L . amazonensis amastigotes are hyper-resistant to the antimicrobial activity of both murine and human MΦs and antagonize iNOS-mediated cytotoxicity in murine cells . Since much of the anti-Leishmania activity of cytokine-activated MΦs was dependent on iNOS , and amastigotes exhibited a survival advantage over promastigotes in these primed cells , we compared the leishmanicidal potential of H2O2 , NO , and ONOO− , cytotoxic molecules produced during the innate MΦ response to Leishmania , against both parasite stages in vitro . Luciferase-expressing promastigotes and amastigotes were treated directly with increasing concentrations of H2O2 , NO , and ONOO− , and luciferase activity at 4 h post-treatment was determined and used as a readout of parasite viability . Amastigotes were significantly more resistant than promastigotes after 4 h of exposure to 125 µM H2O2 ( p>0 . 001 ) ; however , amastigotes became susceptible to higher H2O2 concentrations ( 250 µM-1 mM ) ( Figure 2A ) . Promastigotes exhibited a dose-dependent hyper-susceptibility to the NO-donor spermine NONOate ( sNO ) , with 50% of parasites surviving 4 h after treatment with 1 mM sNO compared to untreated controls ( Figure 2B , open triangles ) . In contrast , 100% of amastigotes survived 1 mM sNO ( Figure 2B , open circles ) . Amastigotes were remarkably resistant to the strong oxidant ONOO− , which was in contrast to the extreme sensitivity of promastigotes ( Figure 2C ) . At 0 . 5 mM concentration of ONOO− , luciferase activity in promastigotes dropped to undetectable levels , while treated amastigotes had comparable luciferase activity to untreated controls . The hyper-resistance of amastigotes to ONOO− suggests that this form of the parasite may possess stage-specific antinitrosative defenses . A truncated TXNPx isoform capable of detoxifying both ROS and RNS and highly expressed by L . chagasi amastigotes was previously described [21] . Based on the enhanced resistance of amastigotes to RNS , we compared the expression level of the 190-amino acid truncated TXNPx1 isoform in L . amazonensis promastigotes and amastigotes . Immunoblots using an antibody against the C-terminal 40 amino acids of the L . pifanoi truncated TXNPx1 isoform [42] revealed that L . amazonensis axenic amastigotes expressed 11-fold higher levels of TXNPx1 than stationary-phase promastigotes ( Figure 3 ) . Moreover , expression of the TXNPx1 was also increased in lesion-derived amastigotes ( Figure 3 ) , indicating that this TXNPx isoform is highly expressed by parasites in developing cutaneous lesions . Given that amastigotes have increased levels of TXNPx1 , and that this antioxidant could mediate parasite protection against MΦ antimicrobial activity , we next determined TXNPx1 protein expression during the course of Leishmania infection in resting and cytokine-activated MΦs . Immunoblots from promastigote-infected MΦs 4 days post-infection indicated that parasite TXNPx1 expression was induced in cytokine-activated MΦs ( Figure 4A , lanes 1 vs . 2 ) . We found a reciprocal relationship between TXNPx1 expression levels and luciferase-expressing parasites in cytokine-activated MΦs , suggesting that conditions that negatively affect parasite survival also increased TXNPx1 expression ( Figure 4A , lanes 1 vs . 2 ) . Given the predicted role of TXNPx1 in RNS detoxification , we then evaluated whether TXNPx1 induction by intracellular parasites was dependent on a functional iNOS hemoprotein . Interestingly , the increase in TXNPx1 expression levels observed in activated MΦs was iNOS-dependent , as judged by the reduced TXNPx1 levels in activated iNOS−/− cells compared to wild-type cells ( Figure 4A , lanes 3 vs . 4 ) . At 48 h post-infection , TXNPx1 expression was also increased in amastigote-infected wild-type cells ( Figure 4B , lanes 2 vs . 3 ) . Moreover , TXNPx1 expression was higher in IFN-γ/LPS-activated MΦs ( Figure 4B , lanes 3 vs . 4 and 5 vs . 6 ) . Similar to the findings for promastigote infection , we found that the increase in TXNPx1 expression in amastigote-infected control and activated MΦs was dependent on iNOS , as indicated by the limited TXNPx1 expression in iNOS−/− cells ( Figure 4B , lanes 7 vs . 8 ) . Due to the induction of TXNPx1 observed in response to host cell activation , we hypothesized that amastigotes derived from cytokine-activated MΦs ( TXNPx1high ) would possess a growth advantage over axenic amastigotes ( TXNPx1low ) in subsequent infection of activated MΦs . To test this hypothesis , we recovered amastigotes from primary cultures [control MΦ , and IFN-γ/LPS-activated MΦs ( with or without the iNOS inhibitor aminoguanidine ) ] and then used them to infect IFN-γ/LPS-activated MΦs ( secondary cultures ) . Cell-derived amastigotes exhibited a statistically significant increase in infectivity compared to axenic amastigotes at 48 h post-infection ( Figure 4C , light grey bar ) . Importantly , TXNPx1high amastigotes harvested from activated MΦs were highly infectious , showing a 3-fold higher parasite load at 48 h post-infection compared to TXNPx1low axenic amastigotes ( Figure 4C , black bar , p<0 . 01 ) . Of note , this enhanced parasite fitness was diminished in the secondary infection if an iNOS inhibitor was included during primary infection ( Figure 4C , dark grey bar ) . Therefore , the parasite response to MΦ RNS is crucial for fostering increased parasite fitness and infectivity . Collectively , these data show that parasites adapt to the harsh intracellular environment of MΦs by increasing antioxidant/antinitrosative defenses , including TXNPx1 induction , which promotes their infectivity and fitness in MΦs . To determine whether amastigotes' hyper-resistance to RNS was attributable to increased expression of TXNPx1 , we generated an L . amazonensis promastigote strain that was stably transfected with the episomal pXG vector harboring the TXNPx1 open-reading frame . Transformed parasites had 3-fold higher TXNPx1 protein expression compared to control parasites carrying the empty pXG vector alone ( Figure S3A ) . Both transformed parasite strains exhibited similar growth kinetics in complete Schneider's medium supplemented with 20% FBS , reaching stationary phase ∼5 days after passage at 1×105 parasites/mL medium ( Figure S3B ) . In addition , the yield of metacyclic promastigotes from both late stationary phase transformant cultures was similar ( Figures S3C ) , indicating that TXNPx1 overexpression did not interfere with metacyclogenesis . Both parasite strains had similar sensitivity to H2O2 at concentrations between 0 . 125 and 2 mM ( data not shown ) . In accordance with the proposed role of TXNPx1 in parasite resistance to nitroxidative stress and increased intracellular survival , promastigotes overexpressing the TXNPx1 isoform were 100-fold more resistant to ONOO−-mediated cytotoxicity than wild-type promastigotes after treatment with 250 µM ONOO− ( Figure 5A , closed symbols ) . Moreover , TXNPx1-overexpressing promastigotes exhibited enhanced intracellular survival in both resting and cytokine-activated MΦs ( Figure 5B ) . Given the intra-MΦ survival advantage exhibited by the TXNPx1-overexpressing parasites , we hypothesized that these parasites could also have increased virulence in murine models . Accordingly , C57BL/6 mice infected with the TXNPx1-overexpressing promastigotes had enhanced lesion development compared to mice infected with control promastigotes , and this elevated pathogenicity correlated to an increase in parasite growth ( Figure 6A and B ) . Using qRTPCR , we confirmed that TXNPx1 mRNA expression was 250-fold higher in cutaneous lesions from TXNPx1-overexpressing parasite-infected mice compared to control lesions ( Figure 6C ) . There was a positive correlation between TXNPx1 mRNA levels and parasite load ( Figure 6D , R2 = 0 . 72 ) . Collectively , these data indicate that increased TXNPx1 expression confers resistance to macrophage microbicidal mechanisms and promotes Leishmania-mediated disease progression .
Leishmania persistence in the phagolysosome is indicative of the parasite's ability to counter host microbicidal activities . Since amastigotes excel in surviving and replicating within mononuclear phagocytes , it is to be expected that this stage possesses exclusive mechanisms to resist and subvert host cell immune responses . Indeed , several studies have shown that amastigotes are much more efficient at disrupting host cell signaling pathways and antimicrobial mechanisms compared to their promastigote counterparts , including blocking NADPH oxidase assembly and suppressing iNOS expression [14] , [15] . The majority of studies have focused on the ability of Leishmania to prevent MΦ activation [43]; however , there is limited knowledge of how the parasites directly defend against antimicrobial products they may encounter within host macrophages . To address some of these issues , we focused our study on parasite interactions with cytokine pre-activated MΦs and present several important findings herein . First , L . amazonensis amastigotes not only resist MΦ microbicidal mechanisms , but also grow robustly in cytokine-activated , antimicrobial-producing MΦs while these cells readily kill promastigotes . Amastigotes are intrinsically more resistant than promastigotes to authentic RNS , implying that they possess developmental stage-enhanced antinitrosative defenses ( Figures 1 and 2 ) . Second , neither promastigotes nor amastigotes apparently interfere with iNOS transcription and NO synthesis in MΦs if the signaling cascades have already been activated prior to the infection ( Figure S2 ) . This view is consistent with the inability of L . amazonensis to block NO production in infected mice [44] . Third , the hyper-resistance of amastigotes to RNS-mediated toxicity is linked to the increased expression of a unique TXNPx isoform ( Figure 3 ) , which belongs to an antioxidant family of proteins known to detoxify several leishmanicidal molecules produced by professional phagocytes [21]–[23] , [45] . We have provided evidence that the TXNPx1 isoform can be induced by host-derived signals during the course of infection ( Figures 4 ) . More importantly , overexpression of the TXNPx1 in promastigotes increases their resistance to direct RNS treatment and MΦ-mediated killing mechanisms during in vitro infection , and also fosters parasite survival and disease progression in an acute murine model of cutaneous leishmaniasis ( Figures 5 and 6 ) . Our findings support the notion that L . amazonensis amastigotes possess distinct resistance strategies that enable them to flourish in the intracellular environment of host cells , causing persistent infection and disease , even in hosts with Th1-like , proinflammatory responses . As expected , promastigotes of L . amazonensis are highly efficient in establishing infection in resting MΦs in vitro , and are capable of replicating in these cells . It was noteworthy that only a small proportion of promastigotes survived the initial killing mechanisms in IFN-γ/LPS-activated , RNS-loaded MΦs , and that the resultant intracellular-transformed amastigote population started to expand around 2 days post-infection . Evidently , constitutively expressed antioxidant molecules in promastigote forms are insufficient to protect them against leishmanicidal molecules elaborated by activated MΦs . We , and others , have previously reported the failure to completely eliminate L . amazonensis promastigotes from classically-activated MΦs in vitro , which is in contrast to promastigotes of L . major and L . braziliensis [31] , [46] , [47]; this is presumably due to the relatively rapid induction of antioxidant capacity by L . amazonensis promastigotes in infected cells [30] . However , due to some potential caveats of using luciferase as a read-out for parasite viability , it is imperative that this assay be coupled with a second parasite quantification method ( such as qRT-PCR ) in order to make the most valid conclusions Our findings of a positive correlation between growth of L . amazonensis amastigotes in cytokine-activated MΦs and expression of a developmentally regulated TXNPx1 are significant . The data indicate that the enhanced resistance of L . amazonensis amastigotes against MΦ antimicrobial activity is closely linked to a developmentally regulated factor capable of directly antagonizing host-derived antimicrobial products . Both L . pifanoi [42] , and L . amazonensis ( Henard et al . , manuscript in preparation ) express naturally truncated TXNPx1 isoforms ( 190 amino acid proteins ) with high homology to the 199-amino acid cytosolic TXNPx except for the C-terminus , which presumably participates in the differential regulation of the TXNPx isoforms . L . chagasi ( a causative agent of visceral leishmaniasis in the New World ) and L . aethiopica ( a causative agent of cutaneous leishmaniasis in the Old World ) also express a truncated TXNPx1 isoform , also referred to as peroxidoxin 1 ( Pxn1 ) , that is preferentially expressed by the amastigote stage [21] , [26] . Interestingly , the L . chagasi TXNPx1 ( Genebank accession # AAG40074 . 1 ) detoxifies both ROS and RNS , while the function of its 199-amino acid TXNPx enzymes ( Genebank accession # AAK69586 . 1 and AAK69587 . 1 ) is limited to ROS [22] . It is likely that the developmentally regulated TXNPx1 is not solely responsible for amastigote resistance to macrophage killing; rather , we suspect that it is part of a complex antioxidant armamentarium , resulting in enhanced resistance of amastigotes to macrophage microbicidal activity . A careful evaluation of the role of the TXNPx1 isoform in vivo will be benefited by the use of a TXNPx1 transgene with mutations in its redox active cysteine residues critical for antioxidant activity . At present , there are no reported studies identifying differentially regulated TXNPx1 isoforms in L . major , L . donovani , L . infantum , or L . tropica , even though their genome sequences are predicted to encode this truncated TXNPx isoform . Ongoing studies in our laboratory are focused on delineating the presence and regulation of other novel antioxidants in several Old and New World Leishmania species . The marked amastigote resistance to host-derived antimicrobial molecules is likely critical for intracellular parasite survival in several different cell types during both acute and chronic phases of disease . In addition to MΦs , Leishmania are phagocytosed by polymorphonuclear neutrophils and dendritic cells at the site of infection [48] , [49] . Indeed , we have recently showed that L . amazonensis amastigotes resist killing by murine neutrophils despite activating the respiratory burst in these cells [50] . Importantly , the enhanced antioxidant/antinitrosative defenses associated with L . amazonensis amastigotes likely enable the parasite to direct its resources to manipulate host cell function rather than repairing cellular damage caused by host-derived antimicrobials . Intracellular parasite killing is required for antigen presentation [51] , [52] , and leishmanial antigen presentation by dendritic cells and MΦs is critical for eliciting protective CD4+ T cell responses that control parasite burden and the disease . Therefore , it is not surprising that L . amazonensis-infected dendritic cells exhibit decreased activation and antigen presentation capabilities compared to L . major- and L . braziliensis-infected cells [53] , [54] , likely due to the superior host antimicrobial resistance of L . amazonensis compared to other Leishmania species [30] . In the future , it will be interesting to examine a correlation between the antioxidant capacity of intracellular parasites and their ability to modulate host cell functions . In a broader context , the findings from this study have several important clinical implications . For example , IFN-γ/TNF-α and RNS production can be detected during chronic human leishmaniasis [55] , [56] , and the development of drug resistance in Leishmania is closely linked to the redox biology of the parasite . Several antimony-resistant Leishmania field isolates have had elevated trypanothione levels , and thiol depletion of these strains reestablishes antimony susceptibility [57] . Moreover , the overexpression of trypanothione-dependent TXN and TXNPx enzymes has been linked to antimony resistance in cutaneous and visceral leishmaniasis [25] , [29] , [58] , and metastasis in mucocutaneous leishmaniasis [59] . Intriguingly , antimony-resistant parasites exhibit hyper-resistance to direct treatment with nitrogen oxides [60] , and show enhanced intracellular survival within MΦs [25] , [61] . At this stage , our understanding of the intracellular role of the TXNPx1 isoform remains incomplete . Specifically , it is unclear whether the TXNPx1 has additional functions independent of its antioxidant activity , or whether it plays a role in altering host gene expression . Nevertheless , a better understanding of Leishmania antioxidant and antinitrosative defenses may lead to the rational design of novel therapeutics for leishmaniasis . In summary , our findings that L . amazonensis amastigotes are highly resistant to MΦ microbicidal defense and respond to NO congeners generated by iNOS are biologically important . Our results indicate that the amastigote-predominant TXNPx1 defends against the antimicrobial effects of ONOO− , and contributes to the unique resistance mechanisms that foster intracellular parasite survival and disease . Since L . amazonensis can cause progressive lesions in patients with diffuse cutaneous leishmaniasis and in most inbred strains of mice and exhibits increased resistance to iNOS-mediated killing compared to other Leishmania species [30] , L . amazonensis is an excellent model for delineating antinitrosative defense mechanisms important for leishmaniasis and drug resistance .
|
Pathogens of the genus Leishmania are the causative agents of leishmaniasis , a neglected tropical disease responsible for significant morbidity and mortality worldwide . Although it is well accepted that host-derived leishmanicidal molecules mediate resolution of Leishmania infection , some Leishmania species/stages are relatively resistant to host cell antimicrobial activity . These intracellular pathogens have developed evasive strategies to subvert host antimicrobials , and promote pathogen survival within the harsh intracellular environment . However , the underlying mechanisms remain largely uncharacterized . Here , we show that L . amazonensis , an agent of persistent infection in humans and non-healing skin lesions in mice , antagonize macrophage antimicrobial activity . The superb ability of the amastigote form to survive within host cells is related to its increased expression of a tryparedoxin peroxidase isoform that confers resistance to the cytotoxicity of host-derived antimicrobial molecules . Parasites induce higher expression of the TXNPx in response to iNOS activity during infection of macrophages , indicating that parasites can “sense” the microenvironment of host cells and regulate the expression of relevant virulence factors accordingly . Our investigations are consistent with a model by which Leishmania amastigotes utilize TXNPx to defend against host-derived molecules thereby promoting their intracellular survival and persistent infection .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"host-pathogen",
"interactions",
"biology",
"and",
"life",
"sciences",
"microbiology",
"pathogenesis",
"parasitology"
] |
2014
|
Leishmania amazonensis Amastigotes Highly Express a Tryparedoxin Peroxidase Isoform That Increases Parasite Resistance to Macrophage Antimicrobial Defenses and Fosters Parasite Virulence
|
The development of organs with particular shapes , like wings or flowers , depends on regional activity of transcription factors and signalling molecules . However , the mechanisms that link these molecular activities to the morphogenetic events underlying shape are poorly understood . Here we describe a combination of experimental and computational approaches that address this problem , applying them to a group of genes controlling flower shape in the Snapdragon ( Antirrhinum ) . Four transcription factors are known to play a key role in the control of floral shape and asymmetry in Snapdragon . We use quantitative shape analysis of mutants for these factors to define principal components underlying flower shape variation . We show that each transcription factor has a specific effect on the shape and size of regions within the flower , shifting the position of the flower in shape space . These shifts are further analysed by generating double mutants and lines that express some of the genes ectopically . By integrating these observations with known gene expression patterns and interactions , we arrive at a combinatorial scheme for how regional effects on shape are genetically controlled . We evaluate our scheme by incorporating the proposed interactions into a generative model , where the developing flower is treated as a material sheet that grows according to how genes modify local polarities and growth rates . The petal shapes generated by the model show a good quantitative match with those observed experimentally for each petal in numerous genotypes , thus validating the hypothesised scheme . This article therefore shows how complex shapes can be accounted for by combinatorial effects of transcription factors on regional growth properties . This finding has implications not only for how shapes develop but also for how they may have evolved through tinkering with transcription factors and their targets .
Although major progress has been made in the genetic dissection of organ and appendage development , the process whereby gene activities lead to particular tissue shapes is still poorly understood . For example , wing morphogenesis in Drosophila is one of the best defined developmental systems [1] , yet little is known about how regional gene activities in the imaginal disc are translated into final wing shape [2] . Addressing this problem has not been easy for several reasons . First , genes that modify shape are normally identified through their overall phenotypic effects , making it difficult to establish how particular regions of the tissue are affected . Second , shape is often described in qualitative terms like “rounder” or “more elongated , ” making it difficult to quantify and compare the effects of different gene combinations . Third , we lack modelling frameworks that allow hypotheses for how genes control morphogenesis to be evaluated quantitatively . Here we combine molecular genetic and morphometric approaches to address these issues , using the Snapdragon ( Antirrhinum majus ) flower as a model system . A key advantage of choosing a plant system is that the lack of cell movement means that morphogenesis arises mainly through differential growth . Shape changes can therefore be described in terms of genes modifying rates of growth in particular orientations [3] . So far , this approach has been applied to studying the effects of genes on overall growth rates of an organ [4] . However , it should be possible to extend this principle to the subregions within an organ , thus allowing final shape to be dissected into genetically determined modulations in the local rates and orientations of growth . The Antirrhinum flower is particularly suitable for this approach as specific shapes can be generated through inactivation or over-expression of key transcription factors . Each flower comprises two upper petals ( dorsals ) and three lower petals ( laterals and ventral ) that together form the corolla ( Figure 1A–D ) . The petals are united proximally to form a tube while the distal regions form five lobes . The shapes of the upper and lower petals are precisely matched at the boundary between tube and lobe , termed the rim , so that the overall structure forms a closed mouth hinged at its edges . The distinctive shapes of the upper and lower petals depend on the activities of four dorsoventral genes: CYCLOIDEA ( CYC ) , DICHOTOMA ( DICH ) , RADIALIS ( RAD ) , and DIVARICATA ( DIV ) [5]–[9] . CYC and DICH encode TCP transcription factors that are expressed from an early stage in the dorsal domain of the flower bud . Mutants lacking both CYC and DICH have flowers with all petals resembling the ventral petal of wild type . RAD and DIV encode Myb-like transcription factors . RAD is switched on by CYC and DICH and promotes dorsal identity , while DIV is active in lower petals and promotes ventral identity . DIV is initially expressed throughout the corolla , but RAD is thought to antagonise its activity , preventing DIV from acting in dorsal petals . At later developmental stages , DIV expression becomes restricted to lateral and ventral petals through the action of the dorsally expressed genes . A cis-acting dominant mutant of CYC ( backpetals ) has been characterised in which CYC is ectopically expressed , leading to lower petals acquiring dorsal identity [9] . However , it is unclear whether the phenotype is a result of ectopic expression of CYC and/or its target gene RAD . The changes in shape resulting from inactivation or over-expression of genes may be quantified using morphometric methods . Such methods have been applied previously to genetically controlled shape variations , such as mandible shape in vertebrates , wing shape in Drosophila , and leaf shape in plants [10]–[13] . This approach involves placing landmarks at key positions on the organ , aligning the resulting points , and then using multivariate methods to extract major trends in variation . The advantages of taking a quantitative approach are that average shapes for each genotype can be extracted and the main features under genetic control can be highlighted . Additionally , this approach potentially allows quantitative comparisons to be made between experimentally generated shapes and shapes generated by computational modelling , enabling hypotheses about morphogenesis to be evaluated . Here we show that the genetic control of flower shape can be accounted for by a combination of region-specific effects . We quantify these effects through shape analysis of previously described mutants and of lines in which RAD is over-expressed in a range of genetic backgrounds . The shapes observed for multiple genotypes can be summarised with a scheme in which dorsoventral transcription factors act in combination with gene activities along the proximodistal and mediolateral axes to modulate the length or breadth of each petal region . Morphogenetic hypotheses for how these phenotypic effects might arise were evaluated using a modelling framework in which genes modify local polarities and specified growth rates [14] , [15] . The petal shapes generated by the resulting model show a good quantitative match with those observed experimentally for each petal from 10 different genotypes , thus validating the underlying hypothesis . Our results suggest that evolution of shape involves a process of “tinkering” , through which size and shape of regions is adjusted by piecemeal modification of local growth properties under the control of transcription factors .
As a first step towards evaluating the effects of different genes on organ shape , the corolla was subdivided into several regions along its proximodistal axis . Most proximal is a continuous cylinder of tissue , the proximal tube . Beyond this region , the tube tissue extends to form the upper and lower palate ( Figure 1E–H ) . The palate ends distally with a boundary called the rim , which acts as a line of transition between the tube and the lobes . The proximal region of the lobes comprises the lip , over which the lobes of adjacent petals are united ( yellow dotted lines in Figure 1E–H ) . The lip is greatly reduced at the junction between the dorsal and lateral lobes , creating a hinge that allows the corolla to be opened by pollinators . The lobes are separate over the remaining distal region of the lobes . To quantify the effects of dorsoventral genes on shape , the outline and size of the various regions of the corolla were captured . First , the 3-D structure of the flower was converted into a series of 2-D shapes . To achieve this conversion , the upper and lower sections of the corolla were separated by making cuts along the junction between lateral and ventral petals . The resulting petal sections were then flattened ( Figure 1I–L ) . Second , the outlines of the regions for each petal were captured using a series of landmarks . Eight primary landmarks ( green dots in Figure 1J , L ) were located at recognisable morphological features , such as where the lobes become separate or where the tube rim and petal junctions intersect . Cell type patterns , which vary along the proximodistal axis of the tube , were also used to define primary landmarks for internal boundaries such as those between ventral and lateral petals . In cases where there were no discernable palate or lip regions , the landmarks bounding these regions were overlaid . The remaining 47 secondary landmarks ( yellow dots in Figure 1J , L ) were spaced evenly along the outlines of each region between the primary landmarks . Taken together , the coordinates for the 55 landmarks summarise the shape and size of the regions for each petal . These coordinate values will vary in a correlated manner between petals depending on how the shapes and sizes of the regions are influenced by genotype and petal identity . The main trends or correlations can be captured using Principal Component Analysis ( PCA ) [16] . To implement this procedure , 110 coordinate values ( from 55 landmarks ) were determined for dorsal , lateral , and ventral petals from wild type as well as the various genotypes described below . Dorsal , lateral , and ventral petals were sampled from five different flowers for each genotype . Petal shapes were aligned by translation and rotation ( Procrustes alignment ) . The average position for each landmark gave the mean petal shape and region outlines for the population . The major trends of variation about this mean were then determined by PCA on the covariance . This analysis showed that 94% of the variance in coordinate positions could be captured with four principal components ( PCs ) . PC1 accounts for 56% of the variance and captures variation in palate and lip size ( Figure 2A ) . Increasing the value of PC1 gives longer petals with extended lip and palate regions , while reducing PC1 gives shorter petals with a reduced lip and palate . PC2 accounts for 23% of the variance and captures petal asymmetry ( Figure 2A ) . Increasing the value of PC2 gives asymmetric petals with shorter lip and palate regions and a longer distal lobe on one side , while reducing PC2 gives bilaterally symmetrical petals . PC3 accounts for 11% of the variance and captures variation in distal lobe size: increasing the value of PC3 gives a smaller distal lobe , while reducing PC3 gives a larger distal lobe . PC4 accounts for 4% of the variance , with an increase in the PC4 value giving a petal that twists in one direction and a decrease giving a petal twisting the opposite way . To determine the contribution of each PC to the specification of petal shapes , average PC values for wild-type dorsal , lateral , and ventral petals were determined and then used to reconstruct the petal shapes ( Table S1 , Figure 2B ) . If all four PC values were used for reconstruction , the resulting shapes closely resemble the observed shapes ( compare top row of Figure 2B with Figure 1J , L ) . This result is expected because these four PCs capture 94% of the variance in petal shape . A good match was also obtained using just PC1 and PC2 , showing that these two PCs are sufficient to capture the main features of the regional shapes . This finding allowed the main shape variations to be represented within a 2-D space that has PC1 and PC2 as its axes . This space will be referred to as the DorsoVentral ( DV ) space ( Figure 2C ) . Each petal sample corresponds to a point in DV space . The origin of DV space , where all PC values are set to 0 , corresponds to the mean petal shape . Samples of the same petal type ( e . g . , dorsal ) form a cloud of points clustered around the mean for that petal type ( Figure 2C ) . The dorsal and lateral clouds are near each other but well separated from the ventral cloud . This clustering reflects the similarity in overall shape and asymmetry of the dorsal and lateral petals and the difference in shape and symmetry of the ventral petals . To determine the effect of the four dorsoventral genes on the ventrally positioned petal , we analysed its shape in several mutant backgrounds . The only dorsoventral gene expressed in the wild-type ventral petal is DIV . The ventral petal of the div mutant therefore expresses no dorsoventral genes and can be considered to represent a ground state . Relative to the wild-type ventral petal , that of div has a reduced palate , is wider , and is not bent back at the rim ( Figure 3B ) . The reduced palate corresponds to a lower value of PC1 ( PC1≈0 ) . The div mutant is therefore shifted to the left in DV space relative to the wild-type ventral petal ( Figure 3K , arrowed ) . The position of the div ventral ground state will be shown in all further DV spaces as a common point of reference . In wild type , expression of DIV in the ventral petal throughout development leads to a longer palate and narrower petal than the ground state . Additionally , the wild-type ventral petal bends back at the rim . These observations indicate that DIV acts to increase palate length , reduce petal width , and promote bending back at the rim . CYC , DICH , and RAD are not expressed in the lower corolla section , so we would not expect these genes to have much effect on ventral petal shape . Consistent with this expectation , the shapes of the cyc dich and rad mutant ventral petals are similar to wild type ( Figure 3C , D ) and map to similar positions in DV space ( Figure 3K ) . In contrast , the ventral petal of backpetals is markedly different from wild type , showing a reduced lip ( Figure 3E ) . The reduced lip size correlates with a leftward shift in DV space ( Figure 3K ) . Additionally , the distal lobe region of backpetals is larger than wild type , particularly along its lateral edges ( giving a low value of PC3; Table S1 ) . Also , similar to the ground state , the ventral lobe does not bend back at the rim in backpetals . Backpetals is a semidominant CYC allele that expresses CYC and its downstream target RAD ectopically in the ventral and lateral petals [9] . The effect of backpetals on ventral petal shape may therefore reflect the action of CYC or RAD or the combined action of both genes . To separate the contributions of CYC and RAD , we generated plants that expressed RAD ectopically , by introducing RAD under the control of the 35S promoter . The ventral petals from these transgenic plants should express RAD but not CYC . Three transgenics were obtained , two of which showed strong petal phenotypes ( Figure 4 ) . No phenotypic effects were observed in leaves , even though RAD expression was detected by RT-PCR of the transgenics but not in wild type ( unpublished data ) . The most noticeable effect of ectopically expressing RAD in the ventral petal was reduction of both the lip and palate regions ( Figures 4 and 3F ) . This reduction resulted in the 35S::RAD point cloud mapping to a similar position to backpetals in DV space ( with a low value of PC1 ) ( Figure 3K ) . Also , like backpetals , the 35S::RAD ventral petal lobe does not bend back . Thus , RAD can exert an autonomous effect on petal shape in the absence of CYC . However , the phenotype of 35S::RAD is not identical to that of backpetals . Unlike 35S::RAD , backpetals has a slightly enlarged medial palate and a large distal lobe ( compare Figure 3E , F ) , indicating that CYC acts partly independently of RAD to increase the length of these regions . To explore interactions between the dorsoventral genes further , 35S::RAD was introduced into several mutant backgrounds ( Figure 3G–J ) . Analysis of the ventral petals showed that the tube of 35S::RAD div resembled that of the div ground state , having a reduced palate ( compare Figure 3B with Figure 3G ) . This result is consistent with previous proposals that a major effect of RAD is to antagonise DIV [7] , [8] . Additionally , the 35S::RAD ventral lip is greatly reduced compared to div , and the palate is also further reduced ( the PC1 value for 35S::RAD is much less than for div ) . This finding indicates that RAD acts independently of DIV to reduce lip and palate length . The phenotype of 35S::RAD in ventral petals resembles that of 35S::RAD rad and 35S::RAD cyc dich . This result is expected because RAD , CYC , and DICH are not normally expressed in ventral petals . In a backpetals mutant background , 35S::RAD had little effect on ventral petal shape , also expected as RAD is already expressed ectopically in the backpetals mutant . We next analysed the effect of dorsoventral genes on dorsally positioned petals ( Figure 5 ) . Wild-type dorsal petals express CYC , DICH , and RAD and also DIV at early stages . The main difference between wild-type dorsal petals and the ground state is the increased value of PC2 , reflecting a marked asymmetry in petal shape . This asymmetry involves a reduced lip on one ( lateral ) side of the petal and an extended palate on the other ( dorsal ) side ( Figure 5A ) . Extension of the palate on the dorsal side of the petal most probably reflects DICH activity , as palate asymmetry is not observed in the ventral petal of backpetals ( Figure 3E ) , which only differs from wild-type dorsal petals in not expressing DICH . Reduction of length on the lateral side of the wild-type dorsal petal depends on RAD activity . In the rad mutant , lip length is restored to this side , reducing the degree of petal asymmetry ( Figure 5D ) . The rad dorsal petals remain asymmetric because DICH activity increases palate length on the more dorsal side . In cyc dich mutants the dorsally positioned petals are fully ventralised ( Figure 5C ) . The petals are bilaterally symmetric because they lack both DICH and RAD expression ( activation of RAD depends on CYC and DICH ) . The absence of RAD also leads to ectopic DIV activity in cyc dich dorsal petals ( RAD normally antagonises DIV ) , accounting for the extended palate and higher value of PC1 relative to the ground state ( Figure 5K ) . If RAD is ectopically expressed in cyc dich dorsal petals ( 35S::RAD cyc dich ) , the PC1 value drops below that of the ground state , as lip and palate regions both become reduced ( Figure 5H ) . This result is consistent with RAD reducing lip and palate length and also further reducing palate length by antagonising DIV . The div mutation does not affect dorsal petal development ( Figure 5B ) , presumably because DIV activity is normally blocked in dorsal petals by expression of RAD . Dorsal petal development is also not affected by the backpetals mutation ( Figure 5E ) , as expected because backpetals does not modify gene expression in the dorsal domain . 35S::RAD also had little or no effect on dorsal petals in wild-type , div , or backpetals backgrounds ( Figure 5G , J ) . Again this result was expected because the endogenous RAD gene is expressed in dorsal petals . 35S::RAD rad dorsal petals have a wild-type phenotype , showing that the transgene complements rad in dorsal regions . This result demonstrates that the shape of the wild-type dorsal petal does not depend on spatial regulation of RAD expression within the dorsal petal . We next analysed laterally positioned petals in various genetic backgrounds . Similar to the wild-type dorsal petal , each wild-type lateral petal is asymmetric with a reduced lip and palate on one ( lateral ) side and extended lip and palate on its other ( ventral ) side ( Figure 6A ) . This morphology places lateral petals in a similar position to dorsal petals in DV space . However , in lateral petals asymmetry of the palate depends on DIV rather than DICH . In the div mutant , the palate is shortened on its ventral side , leading to a more symmetric shape ( lower PC2 value , Figure 6B , K ) . The div lateral petals are still asymmetric because lip and palate length is reduced on the more lateral side of the petal . This reduction involves RAD . In rad mutants , the lateral petal becomes bilaterally symmetrical , with extended lip and palate regions ( Figure 6D ) . The extended palate mainly reflects ectopic DIV activity ( DIV is no longer antagonised by RAD ) , while the extended lip reflects lack of RAD activity . As RAD is not normally expressed in the lateral domain , the reduction of lateral lip growth in wild-type lateral petals involves a non-autonomous effect of RAD expression from the adjacent dorsal domain . If RAD is expressed ectopically in the lateral petal , as in 35S::RAD genotypes , the length of the lip and palate regions becomes negligible and the petal bilaterally symmetrical , with a low PC2 value , similar to that of the ground state ( Figure 6F ) . The value of PC1 value drops below the ground state , reflecting RAD antagonising DIV and also reducing lip length ( Figure 6K ) . Lateral petals of 35S::RAD backpetals are bilaterally symmetrical , like 35S::RAD , but have a partially extended medial palate ( Figure 6J ) . This suggests that expressing CYC counteracts the effect of RAD on reducing palate length in medial regions .
Analysis of petal phenotypes in wild-type , mutant , and transgenic backgrounds reveals that the dorsoventral genes have several region-specific effects on shape . These effects on local shape can be accounted for by a scheme in which the dorsoventral genes interact combinatorially with a pattern of gene activities along the proximodistal and mediolateral axes ( Figure 7 ) . Candidate genes for the proximodistal gene activities are the LIP1 and LIP2 genes , which encode AP2-like transcription factors that increase palate and lip length [17] , and CIN , which encodes a TCP transcription factor that increases lip length [18] . These genes may play an equivalent role to proximodistal systems involved in animal limb development [19] . Less is known about mediolateral systems in plants [20] , although a notable feature in our scheme is that it involves graded changes , allowing lengths to be increased or decreased smoothly . This pattern may be similar to the way graded mediolateral information is provided by Dpp during Drosophila wing development [21] , [22] . The scheme also involves graded effects for RAD activity , which spreads non-autonomously from the dorsal into the lateral domain to restrict DIV function . This spread may reflect direct movement of the RAD protein , as described for other small plant Myb proteins [23] , or more indirect spreading mediated by signalling molecules . Although the scheme in Figure 7 can account for the observed phenotypes through combinatorial effects on the shape and size of regions , it does not define the morphogenetic processes through which shapes are generated . To generate phenotypic outcomes , such as an increase or decrease in length of a petal region , genes presumably modify rates of growth along particular orientations within the region as it develops . However , predicting the consequences of particular hypotheses for growth control can be difficult for several reasons . One is that local orientations may become deformed through differential growth , dynamically modifying the principal orientations in which a region grows . Secondly , the extent to which a region grows may be mechanically constrained by neighbouring regions; so specified growth need not be the same as resultant growth . To address these issues , a computational modelling approach for growing tissues , called the GPT-framework ( Growing Polarised Tissue framework ) , was used to determine the consequences of particular hypotheses [24] . The petal was modelled as a growing material sheet of tissue that can deform in 3-D , incorporating the combinatorial interactions described in Figure 7 [14] . Dorsoventral genes such as CYC and DICH were assumed to be expressed uniformly throughout development within their domains . According to the GPT-framework , genes influence shape by modifying tissue polarity and specified rates of growth ( rates of extension along axes defined by the local polarity ) . For example , the combination DIV·PAL increases palate length by promoting specified growth parallel to the local polarity . Tissue polarity is established through three organisers ( proximal , central , and distal ) , from which polarity signals propagate through the tissue . The activity of these organisers is also influenced by dorsoventral genes [14] . Figure 8 shows the output from the growth model for wild type , from the starting shape of a small lobed cylinder of tissue ( Figure 8A , B ) through to the final shape ( Figure 8C , D ) . To test the hypotheses underlying the computer model , the various genotypes described in this article were generated by setting the relevant gene activity in the model to 0 ( null mutants ) or to 1 everywhere ( over-expression lines ) . The resulting corollas showed a good qualitative match to observed flowers ( Figure 8G–O ) . To give a more quantitative comparison , petals from each model corolla were computationally flattened ( e . g . , Figure 8F ) and their outlines processed in the same way as the observed petal data . The PC values from the model were then compared to the PC values observed experimentally for the corresponding genotype and petal ( Table S1; Figure 8P–S ) . As can be seen in Figure 8P , Q , there is a strong correlation between model output and observational data for PC1 ( R2 = 0 . 87 , p<0 . 0001 ) and PC2 ( R2 = 0 . 91 , p<0 . 0001 ) . This result shows that the model captures the main relationships between genes and shape for each petal and thus provides quantitative validation of the proposed combinatorial interactions between the dorsoventral genes proposed in Figure 7 . Values for PC3 also show a significant correlation between observed and modelled ( R2 = 0 . 56 , p<0 . 0001; Figure 8R ) , suggesting that the model also captures this aspect of petal shape variation . However , PC4 showed little correlation ( R2 = 0 . 04 , p = 0 . 28; Figure 8S ) , which is not surprising because this PC captures only minor shape variations . In the growth model , each dorsoventral gene has several region-specific effects on rates or orientations of growth . This hypothesis is consistent with these genes encoding transcription factors that act in combination with other factors to influence a variety of target genes . These interactions may have been elaborated during the evolution of the Antirrhinum lineage , leading to the formation of a corolla with a closed mouth , hinged at its edges . Such evolutionary tinkering [25] would have included promotion of dorsal and ventral palate growth , by DICH and DIV , respectively , repression of lip growth at the lateral petal boundaries by RAD to create a hinge , and promotion of tissue polarity organisers at particular locations . Thus , the close match between upper and lower petals depends on a history of multiple regional modifications . Similar principles may underlie the close match between the upper and lower jaws of vertebrates , illustrated by mutants in which the lower jaw protrudes or recedes [26]–[28] . The evolution of matched tissue shapes can be compared to the way protein domains may evolve to match each other [29] . In both cases shape-matching arises through tinkering , involving either a sequence of adjustments in regional growth properties and polarities as described here or a series of modifications to protein shape through piecemeal amino acid changes .
Plants of JI 7 ( wild type ) , JI 98 ( wild type ) , JI 726 ( rad-726 ) , JI 609 ( rad-609 ) , JI 721 ( cyc-721 ) , JI 608 ( cyc-608 ) , JI 705 ( backpetals-705 ) , JI 13 ( div-35 [5] ) , and JI 718 ( cyc-608 dich-719 ) were grown in the greenhouse as described previously [30] and recurrently crossed with 35S::RAD transgenic Antirrhinum majus lines . Stocks JI 7 and JI 98 were used as the standard wild type for comparison with the mutants . The 35S::RAD construct was cut from a pGREEN0029 [31] vector and transformed into a binary vector pBIN 19 [32] . This expression vector was transformed into Agrobacterium strain GV3101 and used to transform Antirrhium majus as described by [33] . Three kanamycin resistant shoots were obtained and analysed by PCR using a set of primers for the kanamycin resistance gene ( Neomycin phosphotransferase II ) , 5′-GATGGATTGCACGCAGGTTC-3′ and 5′-GTGGTCGAATG GGC AGGTAG-3′ . A strong phenotype 35S::RAD transgenic line and a weak phenotype 35S::RAD transgenic line were crossed with each of the mutants listed above . The back-crossed plants were screened on MS medium containing 50 mg/l kanamycin and genotyped using the primer sets described below . Genotyping of mutant alleles was performed by PCR using combinations of gene-specific or transposon-specific primers . Primers were 5′-aggttttatgcgacgaattttg3′ and 5′-aggttttatgcgacgaattttg-3′ for rad-726; 5′-atgtttgggaagaacacata-3′ and 5′-ctaattgatgaacttgtgct-3′ for cyc-721; 5′aggttctgactatctgcgcc-3′ and 5′-gtccagtcctttgtcacgtg-3′ for backpetals-705; 5′-atggcttcgactcgtggttc-3′ and 5′-taaggaagcttcgggtccgg-3′ for rad-609; 5′-atgtttgggaagaacacata-3′ and 5′-gtgacccatgcactcttgg-3′ for cyc-608; and 5′-gggtgttccttggacagaag-3′ and 5′-tcatgcgttcggaaagtgaag-3′ for div-35 . The div mutant allele was detected by sequencing PCR products . To detect RAD and transgene expression , total RNA was extracted from young leaves using an RNeasy Plant Mini Kit ( Qiagen , UK ) . First-strand cDNA was synthesised using the SuperScript III First-Strand Synthesis System for RT-PCR ( Invitrogen ) on 5 µg of total RNA treated with a TURBO DNA-free kit ( Ambion ) . RT-PCR was carried out using specific primer sets: 5′-atggcttcgactcgtggttct-3′ and 5′-gaattttgagatttctgaacc-3′ for RAD expression; 5′-agatggattgcacgcaggttc-3′ and 5′-gtggtcgaatgggcaggtag-3′ for NPTII expression; and 5′-attggtgctgaggttgaga-3′ and 5′-acaactgactccagcaaacg-3′ for ubiquitin expression . PCR was performed for 4 min at 94°C and then 30 cycles consisting of 40 s at 94°C , 40 s at 61°C and 60 s at 72°C , followed by 10 min at 72°C . Flower samples were collected from eight individual plants each from mutant and transgenic lines , when flowers were fully opened . Each flower was dissected by cutting in a proximodistal direction along the tube conjunction of dorsal and lateral petals , using a razor . The upper petals ( including two dorsal petals ) and lower petals ( including two lateral petals and a ventral petal ) were flattened by gluing onto paper and photographed using a Nikon Coolpix 995 digital camera . All images were normalised to 4000 pixels/cm2 using an ImagePrep tool written in Matlab . Fifty-five landmarks ( eight primary landmarks and 47 secondary landmarks ) were fitted to each of the dorsal , lateral , and ventral petals ( Figure 1J , L ) to build the shape model using the AAMToolbox ( http://fizz . cmp . uea . ac . uk/wiki/DArT_Toolshed/index . php/Main_Page ) in Matlab ( version: 7 . 2 ) , as described in [12] . A statistical PCA model of flower petal shape and size was generated from the petal point models of the mutant and transgenic plant dataset , projected to a morphospace defined by PC1 and PC2 .
|
A major challenge in developmental biology is to understand how patterns of gene activity are translated into complex three-dimensional forms , like hearts , wings , or flowers . Addressing this problem has not been easy , partly because of the difficulties in quantifying the effects of genes on shape and also because we lack frameworks that allow hypotheses about underlying mechanisms to be evaluated . Here we address this issue through a combination of experimental and computational approaches , using the Snapdragon flower as a model system . By quantifying the shapes of these flowers in a range of mutants with reduced or increased activity of particular genes , we show how the complex floral shape depends on the way genes act in combination in each petal region . The proposed interactions were tested by incorporating them into a computational model of the growing flower . Quantitative comparisons reveal a good agreement between the shapes generated by the model and those observed experimentally , confirming our underlying hypothesis . The Snapdragon flower , with its tightly fitting upper and lower petals , has evolved as a specialised mechanism for targeting pollinators . Our article shows how the development and evolution of such forms may have arisen by natural tinkering with the local effects of genes on growth .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology/organogenesis",
"developmental",
"biology/plant",
"growth",
"and",
"development"
] |
2010
|
Quantitative Control of Organ Shape by Combinatorial Gene Activity
|
Patients with New World cutaneous leishmaniasis ( NWCL ) caused by Leishmania Viannia are treated with parenteral sodium stibogluconate ( SbV ) to reduce the risk of development of mucocutanous leishmaniasis . Our centre manages patients with NWCL on an outpatient-basis . This study was conducted to assess the safety and efficacy of this approach . We reviewed records of 67 consecutive NWCL patients , aged 17–61 years , treated as day-cases with 20 mg/kg/day SbV for up to 28 days at our UK centre . Data had been collected in a standardised format at the time of treatment using a care-record tool . Patients reported adverse-effects daily using a structured questionnaire . Blood tests and electrocardiograms were performed twice weekly to monitor for toxicity . Parenteral SbV treatment was associated with an early , significant suppression of mean lymphocyte and platelet counts . By day four of treatment , lymphocytes reduced by 0 . 53×109/L ( CI 0 . 29×109/L to 0 . 76×109/L , p<0 . 001 ) , and platelets by 31 , 000/µL ( CI 16 , 000/µL to 46 , 000/µL , p<0 . 001 ) . SbV was further associated with significant elevation of serum alanine transaminase concentrations , with a mean peak rise of 107 iu/L by day 13 ( CI 52 iu/L to 161 iu/L , p<0 . 001 ) . These disturbances were temporary and did not result in adverse clinical events . Patient-described symptoms were cumulative and at three weeks of treatment , 59 . 6% of patients experienced myalgia and 29 . 8% malaise . Treatment adherence and clinical outcomes were comparable to inpatient treatment studies . A total of 1407 individual doses of SbV resulted in only 26 nights' hospital admission , a saving of 1381 bed-days compared to inpatient treatment . In specialist centres , NWCL patients aged below 65 years and without co-morbidities can be safely and effectively treated without hospital admission . This reduces the cost of treatment , and is much preferred by patients . Twice weekly blood and electrocardiographic monitoring may be surplus to requirement in clinically well , low-risk patients .
Cutaneous leishmaniasis ( CL ) develops after inoculation of the skin with protozoan parasites of the genus Leishmania , transmitted by the bite of phlebotomine sandflies . Cutaneous nodules form and then ulcerate , typically within weeks of initial infection . The annual worldwide incidence of CL is estimated at 1–1 . 5 million cases , with over 90% of cases occurring in the Middle East and South America [1] . CL may be classified by the geographical origin of the parasite , and New World cutaneous leishmaniasis ( NWCL ) describes CL endemic to Central and South America , most frequently caused by L . viannia and L . mexicana complex . Infection with L . viannia can be complicated by mucocutanoeus leishmaniasis ( MCL ) , when parasites spread to the nasopharyngeal mucosa and produce local destruction [2] . Among indigenous people in rural Bolivia , untreated CL was estimated to progress to MCL in an estimated 5–20% of patients [3] . At the Hospital for Tropical Diseases , London , UK ( HTD ) between 1995 and 2003 we saw 81 cases of active or healed L . viannia NWCL , six of which were complicated by MCL [4] . MCL does not heal spontaneously , is disfiguring and can be fatal . It is important to identify the subgenus of the infecting organism of a NWCL lesion , as parenteral treatment is required in all cases of L . viannia to reduce the risk of developing MCL . Parenteral therapy is also recommended when it is not possible to identify the NWCL species or in ‘complicated’ L . mexicana complex infections ( where there are multiple or large lesions , lesions involve the face or a joint , or there is involvement of the lymphatic system ) , to aid primary lesion healing [5] . Non-complicated , non-L . viannia lesions ( both from the New World and Old World ) should self-heal without dissemination . However , parenteral , intra-muscular or intra-lesional pentavalent antimonials are still recommended as first line treatment in this instance , to accelerate cure and to reduce scarring [2] . The World Health Organisation and the Centres for Disease Control recommend a course of 20 mg/kg/day of sodium stibogluconate ( SbV ) for 20 days to treat NWCL caused by L . viannia [6]–[12] . The administration of SbV is associated with transient toxicities , which have been previously characterized . However , many of the reports on SbV toxicity were published more than 20 years ago or have limitations . In 2011 , Oliveira et al published the first systematic review on the adverse effects of treatments for NWCL and approved only 65 studies on pentavalent antimonials for NWCL as suitable for inclusion , with a total of only 937 patients treated in clinical trials . Furthermore , disparity between treatment regimens employed by different centres , the variability in the antimonium content in each batch of drug produced and a paucity of quantifiable data meant few firm conclusions could be drawn [13] . Patient-described side effects are cumulative and dose-related . A 1998 study of 96 military personnel with NWCL treated with the standard regime reported mylagia in 55% , fatigue in 39% , abdominal pain in 29% and nausea and vomiting in 27% [14] . SbV administration is associated with bone marrow suppression . Hepburn described transient thrombocytopaenia complicating a case of NWCL after standard SbV treatment [15] . Wortmann et al published a prospective study in 1997 of eight NWCL patients in which the white cell and lymphocyte counts were demonstrated to have decreased on day seven of standard SbV treatment [16] . There are case reports of cutaneous and meningitic varicella zoster virus infection during treatment , which have been attributed to SbV-induced leucopenia [16]–[18] . Standard SbV treatment for NWCL is known to derange liver function . Lawn et al demonstrated that in 65 patients , 85% developed a rise in serum hepatic transaminases concentrations . Levels declined after two weeks , despite ongoing treatment [19] . In a study of six patients , Hepburn et al demonstrated acute hepatocellular damage and a fall in the functional metabolic capacity of the liver ( demonstrated by a reduction in caffeine clearance ) that was rapidly reversible on stopping treatment [20] . There are no published reports of fulminant hepatic failure secondary to SbV treatment for NWCL . Elevated serum amylase levels are very commonly associated with SbV administration . Gasser et al showed 48 out of 49 patients receiving standard SbV had elevated amylase levels [21] . However , clinical pancreatitis is rare and severe or fatal pancreatitis extremely rare [22] . Electrocardiogram changes have been described to occur in 50% of patients receiving standard parenteral SbV treatment for NWCL , although cardiac toxicity with tachyarrhythmias is most likely limited to patients with pre-existing cardiovascular pathology [13] , [19] , [23]–[25] . Serious adverse events due to SbV are rare , and deaths are very rare . When serious events do occur , they are most often due to cardiac arrhythmias or pancreatitis . Further data on the toxicity of SbV are available from clinical trials where it has been used to treat visceral leishmaniasis ( VL ) in African countries . Higher rates of toxicity are seen with higher doses of parenteral SbV . Bryceson et al attempted to treat ten cases of VL in Kenya with 60 mg/kg/day of SbV , administered in three divided doses , but this regimen was modified or abandoned in six of the cases due to suspected toxicity [26] . However , it is inappropriate to compare treatment data for patients with systemic VL in Africa and patients with cutaneous-limited disease in the UK . In NWCL-endemic countries , centres may both be familiar with managing the condition and have limited resources and therefore parenteral SbV treatment is usually ambulatory , with hospitalisation often only restricted to a small percentage of patients with co-morbidities or experiencing complications [13] . This is contrary to many centres treating imported NWCL in non-endemic , resource-rich settings , where many clinicians follow the recommendations of Davidson , who advocated inpatient treatment , and monitoring with twice-weekly blood tests and electrocardiograms . Davidson recommended suspending treatment for 1–2 days when liver enzymes reached more than ten times the normal level , amylase three times the normal level , the QTc interval is prolonged to greater than 500 ms or in the event of a cardiac arrhythmia . Davidson's recommendations were based on the limited amount of data available on SbV toxicity and were therefore not strictly evidence-based , giving arbitrary thresholds for treatment suspension [23] . In contrast , in the 2003 study , Lawn et al found SbV toxicity to be of little clinical significance [19] . Therefore since 2003 , treatment for NWCL at HTD is now provided in an outpatient setting , provided that the patient is in good general health , less than 65 years old , and with no known renal , hepatic or cardiac disease . Outpatient care was standardised with the development of a protocol-driven Integrated Care Pathway ( ICP ) . A copy of the HTD ICP can be found in the Supporting Information ( Protocol S1 ) . A prospectively maintained patient database of patients with NWCL has been kept at our centre since the introduction of the ICP , and it is from this database that this study has been conducted . Since 2003 , the HTD has treated up to 20 cases a year of NWCL with parenteral SbV . Cases are identified either when patients self-present to the HTD returned travellers' walk-in clinic or are referred to our centre by general practices , dermatology clinics and military hospitals from across the UK . This retrospective survey had three aims . The first was to determine the toxicity of SbV , when administered at 20 mg/kg/day for 3–4 weeks . This was done both by examining liver , renal and bone marrow function blood tests taken throughout the course of treatment , and by examining the experienced effects prospectively reported by the patients in a daily structured questionnaire . Patients were asked at every treatment if they had symptoms of nausea , myalgia , skin rashes , abdominal pain and/or malaise . The second aim of this study was to determine adherence to the treatment protocol outlined in the ICP by both patients and staff . The ICP recording tool was examined to see whether patients observed the requirement to attend daily for drug administration . The third aim was to determine the efficacy of delivering treatment on an outpatient basis in British patients . There may be differences in responses to treatment and experienced effects by different ethnic group . Prior to the current report , outpatient treatment with parentral SbV in the UK has only been reported in a study on 13 marines [17] .
Our institutional review board ( IRB ) assessed the project and ruled that this project was a retrospective review and a clinical audit and exempted the author group from requiring formal approval . All patient data analysed was anonymised , with patient-identifying information removed , obviating the need for patient consent . Patients were either diagnosed with NWCL after presenting to the HTD walk-in clinic , or were referred to our centre by another hospital or general practitioner . Lesions indicative of NWCL were biopsied for microscopy with Romanovsky type-staining ( Rapid Field's ) , culture with modified Novy-McNeal-Nicolle medium and polymerase chain reaction ( PCR ) for leishmania DNA . Specimens were also examined histologically . A case of NWCL requiring parenteral SbV was defined as a typical lesion ( s ) in a returned traveler from an endemic country in Latin America with confirmation of leishmania infection in at least one laboratory modality and either identification of L . viannia DNA on PCR or identification of a ‘complex’ case with multiple or large lesions or lesions involving the face , a joint , or the lymphatic system . Patients were offered outpatient treatment on the ICP if they meet the following criteria: under 65 years of age , able to attend daily; no pre-existing cardiac or renal condition; normal full blood count , urea and electrolytes and liver function tests; and normal electrocardiogram with corrected QT interval less than 421 milliseconds . We identified all patients who received a course of parenteral SbV for NWCL on the ICP at our centre from April 2003 ( when the ICP was introduced ) until September 2008 from a prospectively maintained clinical research database . Patients travelled to the hospital daily from home or a local hotel . Treatment was for 21 days , which was extended up to 28 days , at the discretion of the attending consultant , if lesion healing was unsatisfactory . Treatment was administered , the lesion was assessed and adverse effects were documented , in accordance with the ICP protocol ( summarised in Table 1 , complete copy of the ICP in Supporting Information ( Protocol S1 ) ) , which was printed in a booklet that also served as a data-recording tool . The nursing staff reviewed and dressed each lesion daily , supplemented by a more detailed weekly inspection . The dose of SbV ( Pentostam® , Wellcome , Brentford , Middlesex , UK ) was 20 mg/kg , with no upper limit , diluted in 100 ml of 0 . 9% saline and administered over 30 minutes via a peripheral venous cannula that was changed every third day . Temperature , pulse and respiratory rates , and blood pressure were measured before , during and after treatment . Nursing staff documented daily if the patient was experiencing myalgia , a skin rash , nausea , malaise or abdominal pain . The team duty doctor was informed if there was any adverse event . Each patient was also medically reviewed every third day , and a full blood count , serum electrolytes , liver function tests and a 12-lead electrocardiogram obtained . Serum amylase was only checked if pancreatitis was clinically suspected . Interruption of treatment was at the discretion of the attending consultant . A retrospective analysis was performed on the case notes , the ICP case recording tool and laboratory results for each patient . The reference ranges of the hospital's laboratories were utilised as the normal ranges for serum concentrations of haemoglobin ( 11 . 5–15 . 5 g/dL ) , lymphocytes ( 1 . 20–3 . 65×109/L ) , eosinophils ( 0 . 0–0 . 4×109/L ) , platelets ( 150 , 000–400 , 000/µL ) , creatinine ( 49–92 µmol/L ) , bilirubin ( 6–23 µmol/L ) , and alanine transaminase ( 10–35 iu/L ) . Patients' symptoms during treatment had been recorded prospectively , using the daily screening questionnaire . The case notes were examined for clinical events suggestive of disordered haemostasis , immune suppression or liver failure . We examined the ICP for the incidence of prolongation of the QTc ( over 420 ms ) , as recorded by the clinicians reviewing the ECGs at time of treatment . The ECGs themselves were not analysed in this study , as this aspect of toxicity has been reported in previous studies [19] , [24] , [25] . Repeated measures analyses-of-variance ( ANOVAs ) was performed on the haematological and biochemical laboratory values measured prior to initiation of treatment and on days four , seven , 10 , 13 , 16 , 19 and 21 of treatment . Pairwise differences were calculated for combinations for each of the dates ( multiple comparison corrected ( Bonferroni ) analysis ) and p-values were calculated . The level of the tests for significance was 0 . 05 . Data were analysed by SPSS for windows version 16 . 0 ( SPSS Inc . , IBM , Somers , New York , USA ) .
Sixty-seven patients ( 50 male ) with NWCL were treated using the ICP between April 2003 and September 2008 . The mean age was 30 years ( range 17–61 years ) . None of the patients had human immunodeficiency virus infection ( routine screening was not performed ) , and none had a significant co-morbid disease . Belize was the most common country of acquisition ( 32/67 , 47 . 7% ) and 25/67 patients were serving in the British Armed Forces and had acquired their CL during military training in the jungles of Belize . Other countries of acquisition were Peru ( 7/67 ) , Bolivia ( 7/67 ) , Guyana ( 5/67 ) , Costa Rica ( 4/67 ) , Brazil ( 3/67 ) , Ecuador ( 3/67 ) , Mexico ( 2/67 ) , Guatemala ( 1/67 ) and Columbia ( 1/67 ) . Of the 67 patients , 51 had infection with L . viannia , 12 had infection with L . mexicana complex , in one patient insufficient protozoal DNA was obtained to determine the infecting species , in two patients PCR was not performed and in one no DNA was found . Systemic therapy was given to the patients with L . mexicana complex infections as there were facial lesions ( 4/12 ) , large or deep lesions ( 4/12 ) , multiple lesions ( 3/12 ) or extensive lymphadenopathy ( 1/12 ) . One third of all patients treated ( 24/67 ) had more than one cutaneous lesion . Lesions were most common on exposed areas ( distal to the elbows , distal to the knees , or over the head and neck ( 85/112 of the total number of lesions ) . There was sporotrichoid spread or lymphadenopathy in lymph nodes adjacent to the skin lesions , palpable on clinical examination , in 19/67 of the cases . No patients had clinical evidence of mucocutaneous infection at the time of treatment . Two of the 67 patients had received prior treatment for their NWCL: one patient had received 21 days of intramuscular SbV while in Argentina , the second had received parenteral SbV , suspended after 12 doses due to deranged liver function , and amphotericin , stopped due to an allergic reaction , at another UK hospital . Table 2 documents the number of daily treatments the patients received . 61/67 ( 91 . 0% ) received at least 20 days treatment . All causes for interruptions to treatment are documented in Table 3 . 12/67 patients ( 17 . 9% ) had disrupted treatment due to toxicity or experienced effects . Patient adherence to the treatment protocol was high , with only two patients failing to attend for six days in total . Only one treatment was omitted due to nursing error . However , in five patients in whom treatments were suspended , staff failed to compensate with additional treatment days at the end of expected treatment . Additional intra-lesional SbV treatment , to assist primary lesion healing , was administered to two patients while they underwent the ICP , and to a further five patients at outpatient follow-up . One of these patients was also given a course of allopurinol to aid healing . These patients are considered presumptive parenteral SbV treatment failures . 19/67 ( 28 . 4% ) patients failed to attend a follow-up appointment following the completion of treatment on the ICP , often , in the case of military personnel , due to redeployment abroad . Of the 48 patients not lost to follow-up , 29 ( 60 . 4% ) and 38 ( 79 . 2% ) patients had lesions that had completely re-epithelialised within two and six months of completing treatment with parenteral SbV respectively ( and had not required additional intra-lesional treatment ) . No patients have reported with re-occurrence of cutaneous disease . Only one of the 44 cases infected with L . viannia developed MCL - 16 months after completing the ICP , the DNA of L . viannia was detected following a nasal biopsy to investigate persistent nasal stuffiness . The mean weight of the female patients was 62 kg and mean daily dose of SbV given was 1230 mg ( range 1000 mg–1600 mg ) . The mean weight for male patients was 78 kg and mean daily dose of SbV given was 1560 mg ( range 1300 mg–1800 mg ) . The study did not identify an obvious association between patients receiving a larger dose of SbV per day ( 31/67 patients received over 1500 mg/SbV/day ) and the toxicity experienced . There was a significant drop in lymphocyte counts between pre-treatment levels and all subsequent lymphocyte levels throughout the course of treatment ( Figure 1 ) . The median lymphocyte count prior to treatment was 2 . 06×109/L , falling to 1 . 57×109/L four days after initiating treatment , a reduction of 0 . 53×109/L , confidence Intervals ( CI ) 0 . 29×109/L to 0 . 76×109/L , p<0 . 001 ( repeated measures ANOVA ) . There was no further significant change in the lymphocyte count , between days seven and 19 , after the initial reduction . Although the median lymphocyte count never dropped below the lower end of normal reference range of 1 . 2×109/L . There were no documented events of opportunistic infections in our patient group . NWCL was not associated with an eosinophilia ( mean eosinophil count pre-treatment 0 . 25×109/L ) , nor did the eosinophil count change during treatment . There was a small but significant drop in the median platelet count in response to treatment ( Figure 2 ) . The median platelet count decreased from 252 , 000/µL before the initiation of treatment to 222 , 000/µL on the fourth day ( a fall of 31 , 000/µL , CI 16 , 000/µL to 46 , 000/µL , p<0 . 001 ) . There was no further significant fall in the platelet counts between days seven to 19 . Throughout treatment , the median platelet count never dropped below the lower end of normal reference range of 150 , 000/µL , and there were no documented events of bleeding . However , two patients became thrombocytopaenic , ( with platelet counts of 32 , 000/µL and 54 , 000/µL respectively ) , both after ten days of treatment . There were no bleeding events , treatment was suspended for seven days in the former patient and continued in the latter , and in both the count recovered to within the normal reference range . There was a small fall in mean haemoglobin concentration by day 19 ( 1 . 0 g/L , CI 0 . 59 to 1 . 43 , p<0 . 001 ) . Serum alanine transaminase ( ALT ) concentrations rose significantly after starting treatment , with considerable individual variability ( Figure 3 ) . The peak median rise was 107 iu/L by day 13 ( CI 52 iu/L to 161 iu/L , p<0 . 001 ) . The levels decreased thereafter , despite ongoing treatment , though this was not significant . In one patient , the ALT rose to 492 iu/L by day 10 , and treatment was suspended for five days . The level had declined to 278 iu/L by day 19 . None of the patients developed a clinically apparent hepatitis . Serum bilirubin and creatinine levels did not rise during treatment . Serum amylase levels were not routinely checked in the patient group . No patient in the group was clinically suspected of having pancreatitis . The peak incidence for reported abdominal pain was on day 18 of treatment ( in five of the 55 patients questioned ) , some of which may have represented symptomatic hyperamylasemia . QTc prolongation ( range 422–524msec ) occurred in nine out of 67 patients , on days four to 19 of treatment . There were no documented events of clinically suspected tachyarrhythmias . The responses to the symptom questionnaire are summarised in Figure 4 . Out of a total of 66 patients for whom there was sufficient data , 63 ( 95 . 5% ) reported at least one adverse reaction . The timing for peak incidences for side effects was: day 15 for skin rashes ( 11/59 of patients whose response was recorded ) , day 18 for abdominal pain ( 5/55 patients ) , day 19 for nausea ( 9/56 patients ) , day 19 for general malaise ( 20/56 patients ) and day 21 for myalgia ( 28/47 patients ) . Although myalgia was common , no patients were suspected to have rhabdomyolysis . Creatine kinase was not routinely checked . For most patients these side effects were mild and transient , and resolved soon after completion of treatment . One patient , who had a history of eczema , had worsening eczema during treatment , which persisted for six months . One patient developed pelvic pain that persisted for four months after treatment finished , but no pelvic pathology was found on investigation . 3/67 patients required one night's hospital admission for drug-related side effects ( one for myalgia , one for vomiting , and one for pyrexia ) . This was out of a total of 1407 doses of drug administered . One patient had their treatment terminated on day 14 for a methicillin-sensitive Staphylococcus aureus bacteraemia secondary to an infected cannula site , which responded to intravenous antibiotics . One patient was admitted for three nights within a week of completing treatment with abdominal pain , vomiting , headache and a temperature that settled with supportive treatment . One patient was admitted for six nights during his treatment as he was of no fixed abode . In total , for the 67 patients treated on the ICP , 1381 bed days were saved .
This case series is a retrospective review and so may be less rigorous than a prospective study , although the use of the ‘Integrated Care Pathway’ protocol in the patients' management and its recording tool has enhanced the quality of our data . As would be expected in any retrospective review , some data was inadequate and in particular we had only limited data on time to lesion healing . The reporting of side effects was subject to bias from both the patient and healthcare provider and ideally the efficacy and toxicity of a drug should be studied in a double-blind randomized control trial as opposed to an observational study . Although our case series of 67 NWCL patients is the largest produced by a UK centre , it remains a relatively small study .
|
Sodium stibogluconate ( SbV ) , a pentavalent antimonial , administered parenterally , is the recommended treatment for South American cutaneous leishmaniasis , caused by Leishmania Viannia , which is a neglected disease that affects many people resident in Central and South America , as well as travellers to the areas . Antimonials have been used for the treatment of leishmaniasis since the 1930s . We report the toxicity experienced by a series of NWCL patients receiving SbV in a resource-rich setting . This study also evaluates administration of the drug to patients without admitting them to hospital . The administration of parenteral SbV was associated with myelosuppression , derangement of markers of liver function and prolongation of the QT interval on electrocardiography , although these effects were not found to be associated with adverse clinical events , and the majority of doses of SbV were administered without cause for hospital admission . Our data shows that parenteral SbV treatment may be provided with reduced monitoring for toxicity than is currently done , and on an outpatient-basis , without endangering safety . Such practice , with reduced demands on local finances and the healthcare workforce , would be desirable in more resource-limited settings .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases"
] |
2012
|
Monitoring Toxicity Associated with Parenteral Sodium Stibogluconate in the Day-Case Management of Returned Travellers with New World Cutaneous Leishmaniasi
|
One of the most important genetic factors known to affect the rate of disease progression in HIV-infected individuals is the genotype at the Class I Human Leukocyte Antigen ( HLA ) locus , which determines the HIV peptides targeted by cytotoxic T-lymphocytes ( CTLs ) . Individuals with HLA-B*57 or B*5801 alleles , for example , target functionally important parts of the Gag protein . Mutants that escape these CTL responses may have lower fitness than the wild-type and can be associated with slower disease progression . Transmission of the escape variant to individuals without these HLA alleles is associated with rapid reversion to wild-type . However , the question of whether infection with an escape mutant offers an advantage to newly infected hosts has not been addressed . Here we investigate the relationship between the genotypes of transmitted viruses and prognostic markers of disease progression and show that infection with HLA-B*57/B*5801 escape mutants is associated with lower viral load and higher CD4+ counts .
Avoidance of host anti-viral responses is a major factor influencing the evolution of HIV genomes . Of particular importance for virus survival , and a major contributor to ongoing HIV-1 diversification worldwide , is continual escape from anti-HIV host cytotoxic T lymphocyte ( CTL ) responses . CTLs can potentially detect many small polypeptide epitope sequences encoded throughout HIV genomes . Evasion of CTL responses involves mutations within and around targeted epitopes that result in the peptide no longer binding to the Class I MHC grove , or non-recognition by the CTL T cell receptor , or interference with peptide processing [1]–[5] . These so-called CTL escape mutations have been associated with increased viral loads and more rapid disease progression [3] , [6]–[8] . However , mutations associated with CTL evasion can also incur significant viral replicative fitness costs and some escape mutations have therefore been associated with decreased viral loads . In the macaque model , for example , in vitro replication of SIVmac239 variants carrying certain CTL escape mutations is impaired relative to SIVmac239 without the mutations [9] , [10] . Fitness costs associated with CTL escape have also been demonstrated in HIV-1 infected humans carrying either the B*57 or B*5801 HLA alleles . CTL escape mutations that frequently arise in these individuals , such as the T242N mutation in the Gag TW10 epitope and the A163X ( X = G , N , D or S ) mutation in the KF11 epitope , have been found to compromise viral replicative capacity [11] , [12] . Because of the fitness costs associated with CTL escape mutations , specific HLA alleles backgrounds of HIV infected individuals have an influence on rates of disease progression . For example , HIV-infected individuals possessing the B*57 , B*5801 and B*27 HLA-alleles tend to take significantly longer to progress to AIDS than individuals without these alleles [13]–[16] . HIV-1 epitopes targeted by these HLA types occur within functionally important protein domains and escape mutations in these domains tend to decrease viral replicative fitness [9] , [11] , [12] , [17]–[19] . Therefore decreased rates of disease progression in people carrying the B*57 , B*5801 and B*27 alleles is at least partially driven by HLA associated virus attenuation . From the virus' perspective , the conflicting demands of replicative fitness on one hand and immune evasion on the other , are best illustrated by the high rates at which certain CTL escape mutations revert to ancestral , presumably replicationally fitter , states following transmission to HLA-mismatched hosts [12] , [20] , [21] . Whenever CTL-escape mutations do not revert following transmission to such hosts it is generally assumed either that the fitness costs of the mutations are negligible [21] , [22] , or that the replicative fitness of escape mutants has been effectively restored by compensatory mutations [11] , [12] , [17] , [21] , [23] , [24] . While much effort has been focused on demonstrating the causal influence of host genetic features on reduced viral replication and decreased rates of disease progression [7] , [25]–[28] , there are a few instances where viral genetic features alone have been identified as the primary cause of slower disease progression . For example , a study dealing with individuals infected with contaminated blood from a common donor determined that long term non-progression was due to transmitted HIV genomes carrying deletions in nef and the terminal repeat region [29] , [30] . Here we describe the identification of two HIV-1 Gag polymorphisms that are associated with low viral loads and high CD4+ counts during both the acute and chronic phases of infection . Although these polymorphisms have been previously identified as attenuating CTL escape mutations in individuals carrying either the HLA-B*57 or B*5801 alleles , we detect these associations in a group of HLA-B*57/5801-negative individuals . We propose that these “attenuating” polymorphisms probably arose during virus passage through HLA-B*57/5801-positive individuals and provide evidence that they have enabled better control of viral replication for up to at least a year following transmission to their current hosts . This is the first demonstration that transmission of HIV variants carrying HLA-associated escape mutations may also afford improved control of virus replication in HLA-mismatched recipients . Our results suggest a dependency between the rate of disease progression in the newly infected host and the genotype of the individual from whom the virus was acquired .
We tested for statistical association between polymorphic amino acid positions and both viral load and CD4+ cell counts . This identified amino acid polymorphisms at two sites in Gag , at HXB2 positions 146 ( n = 9 ) and 242 ( n = 6 ) , that were associated with higher than average CD4+ counts and lower than average viremia ( Table 1 ) . Nine of the 21 study participants were infected with viruses carrying the A146X ( X = P , or S ) polymorphism and the viruses in six of these nine individuals also carried the T242N mutation . Gag amino acid 146 is adjacent to the HLA-B*57/5801 restricted ISW9 epitope and the A146X polymorphism has previously been identified as an epitope processing mutation associated with CTL escape [22] ( Table 1 ) . Similarly , position 242 occurs in the immunodominant TW10 epitope and the T242N polymorphism has also previously been associated with CTL escape in HLA-B*57/5801 positive individuals [21] . We , therefore suspected that viruses carrying either one or both of these two polymorphisms may have been CTL escape variants that had been transmitted from HLA-B*57/5801 positive individuals . Whereas in vitro studies have shown that the A146X mutation does not incur a replicative fitness cost , the T242N is known to decrease viral fitness [11] , [22] . There was only marginal statistical significance ( p = 0 . 0733 ) for an association between the presence of both mutations ( T242N/A146X ) ( n = 6 ) and lower viral loads , however when three additional infections involving viruses carrying the A146X mutation only were included in the analysis ( n = 9 ) , the association was strengthened ( p = 0 . 0275 ) , suggesting that the T242N mutation is not solely responsible for the association ( Table 1 ) . If viruses carrying the T242N and A146X mutations were transmitted from either B*57 or B*5801 positive donors , we hypothesised that selection should be evident at sites within immunodominant B*57 and B*5801 specific epitopes . We analyzed the three B*57 and B*5801 immunodominant epitopes , TW10 , ( TSTLQEQIAW; HXB2 positions 241–249 ) , ISW9 ( ISPRTLNAW; HXB2 positions 147–155 ) and KF11 ( KAFSPEVIPMF; HXB2 positions 162–172 ) for evidence of selection . Comparing sequences from the 21 individuals to the subtype C consensus sequence we calculated the proportions of non-synonymous ( i . e . amino acid-changing ) nucleotide differences that fell within or close to these epitopes ( one flanking amino acid on either side of the epitope was included to allow for possible epitope processing escape mutations ) using the SNAP program ( www . hiv . lanl . gov ) . This analysis indicated that non-synonymous differences from the consensus subtype C sequences were more often associated with B*57 and B*5801 immunodominant epitopes for viruses with the A146X and/or T242N mutations than was the case for viruses without these mutations ( p = 0 . 0010; Figure 1 ) . These results suggests that these sequences had experienced greater selective pressure from the immune response around these immunodominant epitopes and supports our hypothesis that the women from which they were isolated were infected with CTL escape mutants that had arisen in HLA B*57/B*5801 positive individuals . This analysis also revealed additional evidence of selection by B*57/B*5801 restriction . Apart from the potential immune evasion mutations at Gag positions 242 and 146 , three sequences ( CAP088 , CAP228 and CAP255 ) also carried the well-studied A163X mutation in the HLA B*57/B*5801 restricted KF11 epitope ( Figure 2 ) [12] , [31] , [32] . Viruses carrying the T242N and A146X escape mutants from two of the nine individuals ( CAP045 and CAP061 ) , also carried the H219Q compensatory mutation that has been shown to partially restore replicative fitness losses incurred by the T242N mutation [11] , [21] , [33] . The T242N escape mutation is rare in chronically infected HLA-B*57/B*5801-negative individuals and has been known to revert rapidly in these individuals upon transmission from HLA-B*57/B*5801-positive donors [21] . Following limiting cDNA dilution and amplification , Gag sequences from the nine study participants infected with T242N/A146X mutants were analyzed over time to detect reversion mutations ( Figure 2 ) . Reversion of the T242N mutation was observed between six and 24 months post infection in five of the six ( 83% ) individuals initially infected with viruses carrying this mutation ( Figure 2 ) . Reversion of the A146X mutation was only observed in two of the nine ( 22% ) individuals infected with viruses carrying this mutation . These reversions were observed at 16 and 24 months post-infection . Sequences from the two individuals infected with viruses carrying the A163X mutation in the KF11 epitope did not show any reversion of this mutation . To investigate the proportion of T242N and wild-type variants in the six participants infected with T242N mutants , bulk PCR was performed and amplicons from three time points were cloned and sequenced . Although there was complete replacement of the escape mutation ( N242 ) with the consensus amino acid ( T242 ) in three participants ( CAP061 , CAP085 and CAP200 ) , in another participant ( CAP228 ) no reversion was observed ( Figure 3 ) . In the remaining two participants ( CAP088 and CAP225 ) , a mixed viral population consisting of both escape mutants and wild-type variants were detected at the final time point assayed , indicating that complete replacement of the escape mutant with the wild-type variant had not occurred . In CAP225 at 14 . 2 months , the escape mutation was detected in 8/14 clones ( 57% ) , the reversion intermediate , S242 was identified in 4/14 clones ( 29% ) , and T242 occurred in 2/14 clones ( 14% ) . In CAP088 the T242 wild-type was the dominant population member at both 12 . 6 and 18 . 9 months post infection with only , 3/11 and 4/12 of the sampled sequences at these respective timepoints displaying the N242 polymorphism . Reversion of the A146X mutation was only observed in one ( CAP061 ) of the six individuals infected with viruses carrying the T242N mutation . However , although the wild-type A146 polymorphism was observed in 3/10 sampled sequences at 11 . 9 months post infection , it was not detectable amongst ten sequences sampled at 23 . 9 months post infection ( Figure 3 ) . The transience of this reversion indicates that in this participant at least , it may not have provided any substantial fitness advantage . Viral load and CD4+ count dynamics were plotted over time to investigate whether reversion of T242N was associated with either increased viral loads or decreased CD4+ counts ( n = 6 , Figure 4a and b ) . Overall , there was no significant change in geometric mean log viral loads or CD4+ counts between either 6–12 and 12–18 months post-infection , or 12–18 and 18–24 months post infection ( p>0 . 2; Wilcoxon matched pairs test ) ( Table 2 ) . However , one of the six study participants ( CAP085 ) had an increase in log viral load of 1 . 05 and a corresponding decrease in CD4+ count of 209 cells . µl−1 between 12–18 and 18–24 months . The T242 reversion polymorphism was observed in this individual at 6 . 8 months post infection suggesting that the loss of viral control was not concomitant with reversion . These results confirm earlier reports [21] that the T242N mutations revert earlier than other HLA-B*57/B*5801 associated escape mutations and that A146X and other escape mutations persist as “footprints” of prior viral exposure to HLA-B*57/B*5801 alleles . In addition , reversion of T242N mutations to the wild-type consensus sequence does not have an obvious immediate impact on viral load . The 38% frequency of study participants infected with A146X/T242N mutants is higher than the 16 . 5% combined population frequency of HLA-B*57/B*5801 alleles [33] . However , this higher frequency is not inconsistent with all of the individuals infected with A146X/T242N mutants having received these viruses from HLA-B*57/B*5801 positive individuals . Given a sample of 21 individuals from this population we would expect between two and eleven individuals , to be either heterozygous or homozygous for at least one of these two alleles . We nevertheless sought to test whether the observed A146X and T242N mutations had all arisen independently . We analysed our 21 sequences together with 102 other sequences sampled from the same population and within five years of those sampled in our study [16] . We constructed a maximum likelihood tree from these sequences after discarding both nine potentially recombinant sequences and codons corresponding to known HLA-B*57/B*5801 associated escape mutations ( Figure S1 ) . Sequences carrying HLA-B*57/B*5801 associated escape mutations clustered significantly within this tree ( p = 0 . 0157 ) indicating that there is a degree of epidemiological linkage amongst viruses carrying these mutations . This result also reiterates the notion that HLA-B*57/B*5801 associated mutations may persist for extended periods within circulating viruses . Several other HLA alleles besides B*57 and B*5801 have also been associated with improved viral control [34] . It was possible that the reduced viremia and increased CD4+ counts , apparently associated with viruses carrying the A146X/T242N mutations , may have in fact been due to individuals infected with these viruses carrying HLA-alleles that are effective in controlling HIV . Compared to the remainder of the study population , we found no detectable enrichment of any alleles in the nine individuals infected with viruses carrying A146X/T242N mutations ( data not shown ) , indicating that lower viremia and increased CD4+ counts were not obviously associated with over-representation of any one HLA allele . The number of Gag peptides recognised by CD8+ T-cells in ELIspot assays have been shown to be associated with viral control in subtype C HIV-1 infection [16] . To determine if the lower viremia observed in the 9 individuals infected with viruses carrying the putative transmitted CTL escape mutants was associated with the strength or breadth of Gag responses , IFN-gamma ELIspot responses were assessed at 9–15 weeks post-infection ( Table 3 ) . There was no significant difference between the group of individuals infected with the T242N/A146X mutants and that infected with the wild-type virus with respect to ( 1 ) the number of responders , ( 2 ) the magnitude of responses to Gag ( as measured by total sfu/106 PBMC for each Gag pool ) and ( 3 ) the breadth of responses ( as measured by the number of Gag pools recognized ) . Four out of nine individuals infected with T242N/A146X mutants had detectable responses to Gag with three recognizing single peptides ( CAP085 , 225 , 228 ) ( Table 3 ) and the fourth individual recognizing two peptides ( CAP255 ) . Of the 11 individuals infected with the wild-type variant , five responded to Gag , with two individuals recognising single peptides ( CAP008 , 084 ) and three targeting two peptides ( CAP210 , 256 , 257 ) ( Table 3 ) . No responses were detected to Gag fragments carrying the TW10 and ISW9 epitopes . Only one individual ( CAP228 ) showed a response to a peptide overlapping with the KF11 epitope . The optimal epitope within this peptide restricted by the host HLA ( HLA-A*2601 ) is EVIPMFSAL and the observed KF11 A163S escape mutation falls outside this epitope . Together these data support our hypothesis that the T242N and A146X mutations detected in viruses sampled from HLA-B*57/5801 negative individuals are genetic ‘footprints’ of prior passage through HLA-B*57/B*5801 positive donors . Initial identification of these sites was based on a naive scan of all Gag amino acid polymorphisms to identify those associated with high CD4+ counts and low viremia . The proposed mechanism whereby HLA-B*57/5801 individuals achieve good control of HIV replication is unclear , although targeting of the Gag TW10 , ISW9 and KF11 epitopes is thought to contribute to this control [16] , [21] , [34] . Improved control of viral replication in such individuals is due , at least in part , to the fitness costs incurred by the T242N escape mutation in the TW10 epitope . We were therefore interested in determining the specific association of these escape mutations with viremia and CD4+ counts following their transmission to HLA-B*57/B*5801 negative recipients . Clinical data was available from all individuals at 62 days post infection and we compared viral load and CD4+ dynamics up to 15 months post infection in the nine individuals infected with T242N/A146X mutants to those of the rest of the cohort ( Figure 5a and b; Figure S2a and b ) . At all time points the mean log viral load and CD4+ count was lower in the individuals infected with the T242N/A146X mutants . We found that , relative to the rest of the study participants , individuals infected with the T242N/A146X mutants had significantly lower viral loads and higher CD4+ counts at three months post-infection ( median log VL 4 . 53 vs . 5 . 09 , p = 0 . 0077 and median CD4+ count 652 . 0 vs . 460 . 0 , p = 0 . 0129 ) , ( Figure 6a and b ) . At 12 months post infection , these individuals also had significantly lower viral loads and higher CD4+ counts ( median log VL 4 . 26 vs . 4 . 92 , p = 0 . 0275 and median CD4+ count 499 . 0 vs . 322 . 5 , p = 0 . 0172 ) , ( Figure 6c and d ) . This suggests that , in HLA-B*57/B*5801 negative individuals , HIV-1 variants carrying the Gag T242N and A146X mutations tend to be less pathogenic than those which do not carry the mutations . The H219Q mutation is a compensatory mutation reported to partially restore viral fitness [11] . However , we observed that the two individuals infected with H219Q mutant viruses tended to have lower viral loads within the T242N/A146X+ group ( Figure 6a and c ) .
Examination of Gag sequences from acutely infected HLA-B*57/5801 negative women has revealed two polymorphisms , A146X and T242N , that are associated with lower viral loads and higher CD4+ counts in these woman up to a year post infection . As both polymorphisms have been previously identified as HLA-B*57/5801 immune evasion mutations we propose that they are probably genetic footprints of prior virus passage through HLA-B*57/5801 positive individuals . While HLA imprinting of circulating HIV sequences is an established phenomenon [35]–[38] , our demonstration that such imprinting might enable better control of virus replication following transmission to HLA mismatched recipients is entirely novel . While we do not provide definitive evidence that any of the studied viruses has been directly transmitted from HLA-B*57/B*5801 positive individuals , we have detected a strong signal of selection in the Gag HLA-B*57/B*5801 restricted epitope sequences of viruses carrying the A146X and T242N polymorphisms . This is suggestive of at least some of the viruses having been passaged through HLA-B*57/B*5801 positive individuals at some time in the past . Our speculation that the A146X/T242N mutants have been transmitted from either HLA-B*57 or HLA-B*5801 positive individuals is also consistent with previous reports dealing with the persistence of HLA-B*57/B*5801 associated immune evasion mutations . Two Gag mutations , H219X ( X = Q , P or R ) and the A146X polymorphism dealt with here , have been previously identified as relatively stable HLA-B*57/B*5801 genetic imprints on Gag [21] . These mutations appear to be epistatically associated with the T242N mutation but unlike the T242N mutation which reverts following transmission to HLA-B*57/5801-negative hosts , the H219X and A146X mutations are often maintained even in the absence of the selective forces exerted by these alleles [12] , [21] . Whereas the H219X mutation partially alleviates the fitness deficit incurred by the T242N mutation [11] , the A146X mutation has not been associated with any significant fitness loss [22] . Importantly , we detected the H219X mutation in two participants , one associated with a T242N mutation and the other was associated with the A146X mutation . In the latter case the presence of the H219X mutation suggests that this virus may have descended from a T242N mutant . The possibility that most , if not all , of the six women infected with viruses carrying the T242N mutation were directly infected by HLA-B*57/5801-positive individuals is additionally supported by our observation that the mutation reverted in five of the individuals during the study period . It is , however , more uncertain whether viruses carrying the A146X mutation but not the T242N mutation were transmitted directly from HLA-B*57/5801-positive individuals . It cannot be discounted that the A146X mutation in these viruses might be a persistent imprint of a more distant passage through an HLA-B*57/5801 positive individual . While we have detected reversions of this mutation , it has persisted in the viruses infecting seven of the nine individuals studied here for more than two years and is clearly more stable than the T242N mutation . If the mutation had a reversion half-life of four or more years ( as is suggested by our data ) it would not be surprising if some of the A146X mutants we have studied had been serially transmitted two or more times since they first arose . Our detection of significant phylogenetic clustering of gag sequences carrying HLA-B*57/5801 associated escape mutations supports the notion that many of these mutations may persist for epidemiologically significant time periods in HLA-B*57/5801 negative individuals . Reversion of the T242N escape mutation did not result in a concomitant increase in viral load . There is a relationship between viral load during primary infection and viral load set-point [39] and it is possible that these escape mutations provided a long term benefit by reducing viremia during acute infection . It is also possible that , amongst the viruses studied , reversion mutations in gag were not sufficient on their own to fully restore fitness and that there may have been other HLA-B*57/5801 associated escape and compensatory mutations elsewhere in their genomes that impacted on their fitness . Although there was only a marginal correlation between viral load and infection with variants harbouring both T242N and A146X mutations ( p = 0 . 0733 ) , this relationship was much stronger when individuals infected with only the A146X mutation were included in the analysis ( p = 0 . 0275 ) . This supports the existence of a network of B*57/5801 associated mutations that could contribute to viral control . Brockman et . al . [18] recently reported several novel compensatory mutations , associated with the T242N escape mutation , which correlated with higher viral loads . It might have been expected that the two individuals ( CAP045 and CAP061 ) whose viruses had compensatory mutations would have higher viral loads as compared to those infected with the T242N/A146X-carrying viruses [11] , [18] . Viral loads in these individuals were , in fact , lower . While our data suggests that , within the first year of infection , B*57/5801-negative individuals infected with viruses carrying these escape mutations have lower viral loads , the long-term impact on disease progression is unknown . The T-cell responses determined in this study could not explain the differences in viral loads observed for the individuals infected with the escape mutants compared to those infected with the wild-type variants . The number of individuals that responded to Gag did not differ between the two groups and there were no significant differences in the magnitude or breadth of Gag IFN-gamma ELIspot responses between the two groups . However , it is likely that we are underestimating the T-cell responses due to the use of consensus based subtype C reagents compared to autologous peptides . In addition , experimental limitations could also be a contributing factor as the IFN-gamma ELIspot assay does not detect all T-cell responses . We have therefore demonstrated that , during the acute phase of infection at least , individuals who are infected with viruses carrying markers indicative of previous selection in HLA-B*57/5801 positive individuals experience both significantly lower viral loads and higher CD4+ counts than individuals infected with viruses without these markers . These lower initial viral loads and higher CD4+ counts at the onset of infection may slow disease progression in these individuals . The possibility that an interacting network of attenuating mutations may be responsible for the lower viral loads experienced by people infected with viruses passaged through HLA-B*57/5801 positive individuals should be investigated further as the existence of such networks could profoundly influence our understanding of HIV pathogenesis . Current opinion is that first generation CTL based vaccines are likely to be only partially effective . Our study suggests that such vaccines should contain epitopes where escape is associated with a fitness cost to the virus as this might drive the attenuation of viruses in individuals who become infected despite vaccination .
Participants in this study are part of the CAPRISA 002 cohort investigating the role of viral and immunological factors in acute and early HIV-1 infections . The cohort includes high risk HIV negative women monitored monthly for recent HIV-1 infection using two HIV-1 rapid antibody tests and PCR ( Roche Amplicor v1 . 5 ) . HIV-1 infection was confirmed using an enzyme immunoassay ( EIA ) test . Women were enrolled in the present study within 3 months of infection from both the HIV negative cohort , and other seroincidence cohorts in Durban , South Africa . The timing of infection was estimated to be either at the midpoint between the last HIV-1 negative test and the first antibody positive test , or as 14 days where individuals were PCR positive-antibody negative . Samples were collected at enrolment , weekly for three weeks , fortnightly until 3 months , monthly until a year and quarterly thereafter . CD4+ T cells counts were assessed using a FACSCalibur flow cytometer and viral loads were measured using a COBAS AMPLICOR™ HIV-1 Monitor Test v1 . 5 ( Roche Diagnostics ) . Plasma collected in EDTA was stored at −70°C until use . Written informed consent was obtained from all participants . This study received ethical approval from the University of KwaZulu-Natal , University of the Witwatersrand and University of Cape Town . All study participants in the cohort who had reached 12 months postinfection were included in this study , excluding 3 HLA-B*5801-positive individuals . RNA isolated from plasma samples using the Magna-Pure Compact Nucleic Extractor ( Roche ) was reverse transcribed using the Invitrogen Thermoscript Reverse transcription kit ( Invitrogen ) and the primer , Gag D reverse ( 5′-AAT TCC TCC TAT CAT TTT TGG-3′; HXB pos 2382-2402 ) . Limiting dilution nested PCR was carried out by serial end-point dilution of the cDNA [40] . The first round PCR primers were Gag D forward ( 5′-TCT CTA GCA GTG GCG CCC G-3′; HXB pos 626–644 ) and Gag D reverse ( 5′-AAT TCC TCC TAT CAT TTT TGG-3′; HXB pos 2382–2402 ) . The second round PCR primers were Gag A forward ( 5′-CTC TCG ACG CAG GAC TCG GCT T-3′; HXB pos 683–704 ) and Gag C reverse ( 5′-TCT TCT AAT ACT GTA TCA TCT GC-3′; HXB pos 2334–2356 ) . PCR products were either directly sequenced or cloned using the -T Easy vector system ( Promega ) . Sequencing was carried out using an ABI PRISM dye terminator cycle-sequencing kit ( Applied Biosysytems ) and the primers Gag A forward , Gag A reverse ( 5′-ACA TGG GTA TCA CTT CTG GGC T-3′; HXB pos 1282–1303 ) , Gag B forward ( 5′-CCA TAT CAC CTA GAA CTT TGA AT-3′; HXB pos 1226–1246 ) , Gag B reverse ( 5′-CTC CCT GAC ATG CTG TCA TCA T-3′; HXB pos 1825–1846 ) , Gag C forward ( 5′-CCT TGT TGG TCC AAA ATG CGA-3′; HXB pos 1748–1768 ) and Gag C reverse for direct sequencing . For cloned sequences , only the Gag B forward and Gag B reverse primers were used generating p24 gag sequences . Sequences were assembled using the CAPRISA Assembly Pipeline tool ( http://tools . caprisa . org/ ) and aligned using ClustalW ( with default settings [41] ) . High resolution ( four digit ) HLA typing was performed on all participants . DNA was extracted from either PBMCs or granulocytes using the Pel-Freez DNA Isolation kit ( Pel-Freez ) . HLA-A , -B and -C typing was performed by sequencing of exons 2 , 3 and 4 using Atria AlleleSeqr kits ( Abbott ) and Assign-SBT 3 . 5 ( Conexio Genomics ) . Any ambiguities resulting from either polymorphisms outside the sequenced exons or identical heterozygote combinations , were resolved using sequence-specific primers . PBMC were prepared and HIV-1 specific T cell responses were quantified by gamma interferon ( IFN-γ ) Elispot assay [42] . Synthetic overlapping peptides ( 15- to 18-mer peptides overlapping by 10 amino acids ) spanning the entire HIV-1 clade C Gag protein corresponding to the HIV-1 consensus C were used in the assay . T cell responses were derived using either a deconvoluted pool matrix approach or confirmed using individual peptides . The epitopes within peptides showing responses were predicted from the published epitopes on the Los Alamos HIV database ( www . hiv . lanl . gov/content/immunology ) . Phylogenetic trees were constructed using the maximum likelihood method implemented in PHYML [43] ( 100 full maximum likelihood bootstrap replicates , HKY model+gamma with four substitution rates and transition:transversion ratio determined from the data ) . Seven different recombination analysis methods implemented in RDP3 [44] were used , with default settings , to test for the presence of recombination amongst the 21 acute infection sequences and an additional 102 publicly available gag sequences sampled from a matched cohort . Evidence of phylogenetic clustering of viruses carrying particular Gag polymphisms was assessed using a permutation test ( with 10000 iterations ) implemented in RDP3 that is similar to that described in Poss et al . [45] . Wilcoxon rank-sum tests were used to identify amino acid sites ( encoded by the earliest gag sequences determined post infection ) that were associated with low viral loads and high CD4+ counts at 12 months postinfection . These tests compared the median viral loads and CD4+ counts between groups of viruses with the consensus or an alternative amino acid at each site independently ( without correction for multiple testing ) . Fisher's exact test was used to test each HLA allele for enrichment among individuals with either the A146X or T242N mutations and to test for associations between Gag ELIspot responses in the two groups ( with and without T242N/A146X mutations ) . Changes in viral loads were tested using the Wilcoxon matched pairs test . Statistical tests were implemented in the R statistical computing environment [46] and GraphPad Prism 4 . 0 ( GraphPad Software , Inc . ) . Sequence data are available from GenBank under accession numbers EU347404–EU347714 .
|
Following infection with HIV , it is well established that a person's genetic makeup is a major determinant of how quickly they will progress to AIDS . Particularly important is the class I Human leukocyte antigen ( HLA ) gene that is responsible for alerting the immune system to HIV's presence . One of the reasons our immune systems are unable to beat HIV is that the virus can mutate to forms that our HLA genes no longer recognise . However , some people have versions of the HLA gene ( for example HLA-B*57 and HLA-B*5801 ) that are known to force HIV to tolerate mutations that damage its ability to reproduce . Slower HIV reproduction is thought to be one reason that HLA-B*57 and HLA-B*5801 positive people progress to AIDS more slowly than most other HIV infected persons . We report here on a study of HLA-B*57 and HLA-B*5801 negative women in which better control of disease tended to be associated with their being infected with viruses carrying mutations that have been previously shown to reduce replication . These mutations characterise viruses found infecting HLA-B*57 and HLA-B*5801 positive people . This indicates for the first time that HLA-B*57 or HLA-B*5801 negative people that are infected by such reproductively compromised viruses may also experience better survival prospects .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"infectious",
"diseases/hiv",
"infection",
"and",
"aids",
"infectious",
"diseases"
] |
2008
|
Transmission of HIV-1 CTL Escape Variants Provides HLA-Mismatched Recipients with a Survival Advantage
|
Mutations of SLC26A4 are a common cause of human hearing loss associated with enlargement of the vestibular aqueduct . SLC26A4 encodes pendrin , an anion exchanger expressed in a variety of epithelial cells in the cochlea , the vestibular labyrinth and the endolymphatic sac . Slc26a4Δ/Δ mice are devoid of pendrin and develop a severe enlargement of the membranous labyrinth , fail to acquire hearing and balance , and thereby provide a model for the human phenotype . Here , we generated a transgenic mouse line that expresses human SLC26A4 controlled by the promoter of ATP6V1B1 . Crossing this transgene into the Slc26a4Δ/Δ line restored protein expression of pendrin in the endolymphatic sac without inducing detectable expression in the cochlea or the vestibular sensory organs . The transgene prevented abnormal enlargement of the membranous labyrinth , restored a normal endocochlear potential , normal pH gradients between endolymph and perilymph in the cochlea , normal otoconia formation in the vestibular labyrinth and normal sensory functions of hearing and balance . Our study demonstrates that restoration of pendrin to the endolymphatic sac is sufficient to restore normal inner ear function . This finding in conjunction with our previous report that pendrin expression is required for embryonic development but not for the maintenance of hearing opens the prospect that a spatially and temporally limited therapy will restore normal hearing in human patients carrying a variety of mutations of SLC26A4 .
Enlargement of the vestibular aqueduct ( EVA; OMIM #600791 ) is a malformation of the temporal bone that is commonly observed in children with sensorineural hearing loss [1] , [2] , [3] , [4] , [5] . Mutations of SLC26A4 are the most common cause for EVA-associated hearing loss that can either be non-syndromic ( DFNB4; OMIM # 600791 ) or syndromic with enlargement of the thyroid gland ( Pendred syndrome; OMIM #274600 ) . SLC26A4 codes for the anion exchanger pendrin that transports anions such as Cl− , I− and HCO3− [6] , [7] . Although EVA is a malformation of the temporal bone , it is not the cause for hearing loss since no correlation was found between the degree of EVA and the severity of hearing impairment [8] . EVA , however , is an indication of an enlargement of the endolymphatic duct epithelium that was present during embryonic development . Cartilage cells that form in the periphery of the endolymphatic duct epithelium preserve the diameter of the duct in a ‘fossil-like’ record when they give rise to the bone of the vestibular aqueduct . The mature inner ear consists of seven interconnected fluid spaces that house six sensory organs ( Fig . 1 ) : The cochlea for hearing , the utricle and saccule for sensing linear acceleration including gravity , and three ampullae with semicircular canals for sensing angular acceleration in three spatial axes . The seventh fluid compartment is the endolymphatic duct and sac , which is devoid of sensory cells and which is suspected to play a role in fluid homeostasis [9] , [10] . Pendrin is expressed in a variety of epithelial cells that enclose endolymph , which is the luminal fluid of the inner ear ( Fig . 1 ) . Pendrin is expressed in outer sulcus , spiral prominence and spindle-shaped cells in the cochlea , transitional cells in the utricle , saccule and ampullae and mitochondria-rich cells ( synonym: Forkhead-related or FORE cells ) of the endolymphatic sac [11] , [12] . Each cell type represents a small domain in the heterogeneous epithelium that encloses endolymph . The many locations and cell types that express pendrin in a normal inner ear made the goal to restore function through restoration of expression look futile unless some sites of expression were more important than others . The earliest onset of pendrin expression in the murine inner ear occurs in the endolymphatic sac at embryonic day E11 . 5 , which precedes the onset of expression in the cochlea by 3 days , in the saccule and utricle by 4 days , and in the ampullae by 5 days [13] . The expression in the endolymphatic sac surges dramatically at E14 . 5 , a time in development when there is very little pendrin expressed elsewhere in the inner ear [13] . Studies in a mouse model , Slc26a4Δ/Δ , have revealed that loss of pendrin leads to an enlargement of endolymph volume followed by an acidification and a failure to develop normal hearing and balance [13] , [14] . The onset of the enlargement in the cochlea and the endolymphatic sac occurs at E14 . 5 which precedes the onset of the luminal acidification by 1 day in the cochlea and by 3 days in the endolymphatic sac [13] . The enlargement develops in Slc26a4Δ/Δ mice between E14 . 5 and E18 . 5 , which is the phase of rapid growth of the cochlea [4] . The coincidence of the surge in pendrin expression in the endolymphatic sac at E14 . 5 and the onset of the enlargement in Slc26a4Δ/Δ mice points to the importance of pendrin expression in the endolymphatic sac for inner ear fluid homeostasis . We hypothesized that restoration of pendrin expression in the endolymphatic sac would prevent enlargement and permit normal development of the cochlea and the vestibular labyrinth including the acquisition of sensory function . To test this hypothesis , we generated a mouse line that expresses human pendrin SLC26A4 controlled by the promoter of the B1-subunit of the human vacuolar H+ ATPase ( ATP6V1B1 ) and crossed this transgene into the Slc26a4Δ/Δ line to generate mice that lack expression of mouse pendrin but express human pendrin in the endolymphatic sac . No expression of pendrin protein was detected in these mice in the cochlea or the vestibular labyrinth but in mitochondria-rich cells of the endolymphatic sac . Analysis of this mouse model revealed normal hearing and balance function . Our data indicate that the expression of pendrin solely in the endolymphatic sac of the inner ear is sufficient to permit the development of normal hearing and balance .
A transgenic mouse line , referred to as Tg ( B1-hPDS ) Tg/+; Slc26a4+/+ and abbreviated here to Tg ( + ) ;Slc26a4+/+ , was created by the laboratory of Dr . Dominique Eladari ( Paris , France ) [15] . This mouse expresses human SLC26A4 ( formerly named hPDS ) controlled by the promoter of ATP6V1B1 , which codes for the B1-subunit of the vH+ATPase . Transgenic founders were crossed with wild-type C57BL/6× CBA F1 mice and three Tg ( + ) ;Slc26a4+/+ mice were shipped to Kansas State University ( Manhattan , Kansas , USA ) . At Kansas State University , Tg ( + ) ;Slc26a4+/+ mice were crossed with Slc26a4Δ/Δ mice , which are maintained in an isogenic 129S6SvEv background , to generate the desired Tg ( + ) ;Slc26a4Δ/Δ mice in an F2 generation . Expression of SLC26A4 ( human pendrin ) in this mouse was expected to originate solely from the transgene since Exon 8 in the Slc26a4Δ allele was replaced with a neomycin-cassette that introduced a frame-shift [14] . Littermates with the genotype Tg ( − ) ;Slc26a4Δ/Δ served as negative controls . These mice were expected to lack functional pendrin protein expression . Further , littermates with genotypes Tg ( − ) ;Slc26a4Δ/+ , Tg ( + ) ;Slc26a4Δ/+ , and Tg ( + ) ;Slc26a4+/+ served as positive controls . These mice were expected to express murine pendrin with or without augmentation of human pendrin and have normal hearing and balance . Expression of Atp1a1 , Atp6v1b1 , Slc26a4 and SLC26A4 was determined by quantitative RT-PCR and normalized to the expression of 18S rRNA ( Fig . 2 ) . The highest levels of Atp6v1b1 and Slc26a4 mRNA among the different inner ear tissues were found in the endolymphatic sac ( Fig . 2B and C ) . Expression of Slc26a4 was reduced by factors between 6 and 16 in Tg ( + ) ;Slc26a4Δ/Δ mice compared to Tg ( − ) ;Slc26a4Δ/+ mice ( Fig . 2C vs G ) . Expression levels of Atp1a1 and Atp6v1b1 exhibited a similar pattern among inner ear tissues of Tg ( − ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 2A vs E and Fig . 2B vs F ) . Most interesting , expression levels of human SLC26A4 in Tg ( + ) ;Slc26a4Δ/Δ resembled the pattern of mouse Slc26a4 in Tg ( − ) ;Slc26a4Δ/+ mice with the highest levels being expressed in the endolymphatic sac ( Fig . 2C vs H ) . Whether or not expression levels of human SLC26A4 in Tg ( + ) ;Slc26a4Δ/Δ exceeded expression levels of mouse Slc26a4 in Tg ( − ) ;Slc26a4Δ/+ mice remained undetermined , since the efficiency of the reverse transcription of mRNA into cDNA remains generally unknown in quantitative RT-PCR experiments . Taken together , the data demonstrate that the transgene restored pendrin mRNA expression to the endolymphatic sac , the cochlea and the vestibular labyrinth of the inner ear . The ability of the ATP6V1B1 promoter to drive protein expression in different tissues including the cochlea , the vestibular labyrinth and the endolymphatic sac was evaluated in a transgenic mouse line , Tg ( B1-eGFP ) in which the expression of eGFP is controlled by the same 6 . 9 kb promoter of the human ATP6V1B1 gene that drives the expression of human pendrin in Tg ( + ) ;Slc26a4Δ/Δ mice [16] . No expression of eGFP was detected in the cochlea or the vestibular labyrinth of E15 . 5 Tg ( B1-eGFP ) mice , although expression was present in the endolymphatic sac and the kidney ( Fig . 3 ) . These data suggest that the ATP6V1B1 promoter does not drive protein expression in the cochlea or the vestibular labyrinth . Soft tissues of the cochlea and the vestibular labyrinth , exclusive of the endolymph sac , were collected from adult mice by microdissection and pooled into an ‘inner ear’ sample . Crude membrane protein preparations were obtained from these inner ear samples and from kidneys and subjected to gel-electrophoresis and Western blotting . Membrane proteins were obtained from Tg ( + ) ;Slc26a4Δ/Δ mice as well as from Tg ( − ) ;Slc26a4Δ/+ mice , which served as positive controls , and from Tg ( − ) ;Slc26a4Δ/Δ mice , which served as negative controls . Pendrin was detected in the inner ear and kidney of Tg ( − ) ;Slc26a4Δ/+ mice as a ∼110 kDa band ( Fig . 4A ) . Inner ear from Tg ( + ) ;Slc26a4Δ/Δ mice lacked this band . The observation that there was no difference in the pattern of faint bands between inner ears from Tg ( + ) ;Slc26a4Δ/Δ mice and Tg ( − ) ;Slc26a4Δ/Δ mice , which is the negative control , suggests that pendrin was either not detectable or not present . The pendrin band , however , was found in kidney from Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 4B ) , which suggests that the antibody recognizes both mouse and human pendrin . The observation that pendrin was detected at similar levels in descending amounts of kidney proteins isolated from Tg ( − ) ;Slc26a4Δ/+ mice ( Fig . 4A ) and Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 4C ) suggests that the detection threshold for mouse and human pendrin was similar . Whether the antibody differed in the sensitivity between mouse and human pendrin remains unknown , since the relative abundance of mouse pendrin in kidneys of Tg ( − ) ;Slc26a4Δ/+ mice and human pendrin in kidneys of Tg ( + ) ;Slc26a4Δ/Δ mice is not known . The meaning of pendrin being not detectable in the inner ear was evaluated by comparison of the intensity of the pendrin band in inner ear to the intensities in descending amounts of kidney protein ( Fig . 4A ) . This comparison suggests that a ∼5-fold lower amount pendrin should have been detectable in the inner ear . This means that pendrin expression in the inner ear of Tg ( + ) ;Slc26a4Δ/Δ mice is either absent or expressed at a level that does not exceed 20% of the expression level in Tg ( − ) ;Slc26a4Δ/+ mice . Temporal bones from Tg ( + ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4Δ/Δ mice were isolated at E15 . 5 , fixed and injected with white paint ( Fig . 5A and B ) . Most striking is that there was no enlargement of the endolymphatic sac , duct or cochlea in Tg ( + ) ;Slc26a4Δ/Δ mice and that the morphology of Tg ( + ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4Δ/Δ mice was grossly similar . These data demonstrate that the introduction of the transgene rescued the malformation previously described in Slc26a4Δ/Δ mice [10] , [14] . Whole-mounted specimens of the endolymphatic sac were prepared for immunocytochemistry from Tg ( + ) ;Slc26a4Δ/Δ , Tg ( − ) ;Slc26a4Δ/Δ , and Tg ( + ) ;Slc26a4Δ/+ mice ( Fig . 5C–E ) . Most striking is the enlargement and lack of pendrin expression in the endolymphatic sac of Tg ( − ) ;Slc26a4Δ/Δ mice ( Fig . 5E ) and the similarity in size and similarity in pendrin expression between the endolymphatic sac of Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 5C ) and Tg ( + ) ;Slc26a4Δ/+ mice ( Fig . 5D ) . These data demonstrate that the transgene drives pendrin expression in the endolymphatic sac and that the introduction of the transgene rescued the malformation [10] , [14] . Gross morphological examination of inner ears revealed greater similarity between Tg ( + ) ;Slc26a4Δ/Δ mice and Tg ( + ) ;Slc26a4Δ/+ than between Tg ( + ) ;Slc26a4Δ/Δ mice and Tg ( − ) ;Slc26a4Δ/Δ mice ( Fig . 6A–C ) . Cochlear turns in Tg ( + ) ;Slc26a4Δ/Δ mice appeared normal in width and did not show widening of turns or thinning of the otic capsule that was seen in Tg ( − ) ;Slc26a4Δ/Δ and that was previously described in Slc26a4Δ/Δ mice [12] , [17] . Inspection of the oval window revealed ‘glittering’ otoconia in the saccule in Tg ( + ) ;Slc26a4Δ/Δ and Tg ( + ) ;Slc26a4Δ/+ mice in contrast to Tg ( − ) ;Slc26a4Δ/Δ mice where no ‘glittering’ was visible ( Fig . 6D–F ) . Midmodiolar sections of cochlear tissues were prepared for immunocytochemistry from Tg ( + ) ;Slc26a4Δ/Δ mice and positive controls consisting of Tg ( − ) ;Slc26a4Δ/+ or Tg ( + ) ;Slc26a4Δ/+ mice . No evidence for cochlear enlargement was found in Tg ( + ) ;Slc26a4Δ/Δ mice at E16 . 5 ( Fig . 7B ) , P1 ( Fig . S1A ) , P16 ( Fig . 7A; Fig . S1E ) or P18 ( Fig . S1C ) suggesting that the introduction of the transgene rescued the cochlear malformation previously described in Slc26a4Δ/Δ mice , which includes a ∼10-fold enlargement of the cochlea [12] , [14] . No detectable pendrin expression was found in the spiral prominence or outer sulcus epithelium of the cochlea in Tg ( + ) ;Slc26a4Δ/Δ mice although prominent expression was observed in these cells in positive controls ( Fig . 7 and S1 ) . The absence of pendrin in Tg ( + ) ;Slc26a4Δ/Δ mice was observed with two different anti-pendrin antibodies ( Pds #1 and Pds #2 ) . The patterns of pendrin expression in the positive controls , Tg ( − ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4Δ/+ mice , were similar for both antibodies to the pattern previously observed in Slc26a4Δ/+ mice [11] , [12] , [13] . Expression of pendrin was further examined in whole-mounted specimens that encompassed the spiral limbus , organ of Corti and outer sulcus . No detectable pendrin expression was found at age P35 in the spiral limbus of Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 7C ) or Tg ( − ) ;Slc26a4Δ/+ mice ( Fig . 7H ) in contrast to the prominent expression of pendrin in the outer sulcus epithelia of Tg ( − ) ;Slc26a4Δ/+ mice ( Fig . 7J ) . For completeness , it needs to be reported that some punctate staining was found in nerve terminals near inner hair cells of Tg ( + ) ;Slc26a4Δ/Δ ( Fig . 7D ) and Tg ( − ) ;Slc26a4Δ/+ mice ( Fig . 7I ) . The endocochlear potential and the difference in pH between endolymph and perilymph was measured with double-barreled ion selective electrodes in Tg ( − ) ;Slc26a4Δ/Δ mice , Tg ( + ) ;Slc26a4Δ/+ mice , and Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 8 ) . Tg ( − ) ;Slc26a4Δ/Δ mice failed to develop a normal endocochlear potential and the pH of endolymph was lower ( = more acidic ) than in perilymph , as previously reported [18] . In contrast , Tg ( + ) ;Slc26a4Δ/+ mice , as reported for Slc26a4Δ/+ mice [18] , developed a normal endocochlear potential and a normal endolymphatic pH that was higher ( = more alkaline ) than in perilymph . Similar to Tg ( + ) ;Slc26a4Δ/+ mice , Tg ( + ) ;Slc26a4Δ/Δ mice developed a normal endocochlear potential and a normal endolymphatic pH even though no detectable pendrin expression was observed in the cochlear epithelium . These data demonstrate that the introduction of the transgene , which rescued the malformation , also rescued the loss of the endocochlear potential and the loss of normal endolymphatic pH homeostasis . Hearing tests were based on auditory brain stem recordings and thresholds in response to tone bursts of 8 kHz , 16 kHz and 32 kHz . Tests performed in Tg ( + ) ;Slc26a4+/+ , Tg ( − ) ;Slc26a4Δ/Δ and Tg ( + ) ;Slc26a4Δ/Δ mice confirmed profound deafness in Tg ( − ) ;Slc26a4Δ/Δ mice ( Fig . 9B ) consistent with previous findings in Slc26a4Δ/Δ mice [14] , [18] . Waveforms of auditory brain stem recordings as well as thresholds were similar between Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 9A ) and Tg ( + ) ;Slc26a4Δ/+ ( Fig . 9C ) . These findings demonstrate that the introduction of the transgene rescued normal hearing although the cochlea did not express detectable levels of pendrin . We next evaluated whether the rescued hearing phenotype in Tg ( + ) ;Slc26a4Δ/Δ would be stable through at least 3 months of age . Auditory brain stem recordings were performed in Tg ( + ) ;Slc26a4Δ/+ mice ( Fig . 9D–F ) , in Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 9J–L ) , and in Tg ( + ) ;Slc26a4+/+ mice ( Fig . 9G–I ) at 1 , 2 and 3 month of age . Hearing in Tg ( + ) ;Slc26a4Δ/Δ mice at 8 kHz and 16 kHz was stable through 3 months . Thresholds were very similar among individuals and did not differ from Tg ( + ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4+/+ mice . A greater variability in hearing thresholds was observed at 32 kHz ( Fig . 9L ) , with 10 of the 19 Tg ( + ) ;Slc26a4Δ/Δ mice maintaining excellent hearing ( thresholds ≤30 dB at 32 kHz ) and 5 developing a high-frequency hearing loss ( thresholds ≥60 dB at 32 kHz ) . About one half of the Tg ( + ) ;Slc26a4Δ/Δ mice ( 9 of 19 ) developed progressive threshold elevations at 32 kHz with thresholds increasing by ≥10 dB between the monthly measurements . This variability is reflected in the greater error bars at 32 kHz but did not lead to a statistically significant difference between Tg ( + ) ;Slc26a4Δ/Δ and Tg ( + ) ;Slc26a4+/+ mice . Sections and whole-mounted specimens of vestibular tissues were prepared for immunocytochemistry from Tg ( + ) ;Slc26a4Δ/Δ mice and positive controls consisting of Tg ( − ) ;Slc26a4Δ/+ or Tg ( + ) ;Slc26a4Δ/+ mice . No evidence of pendrin expression was found in the three sensory organs in Tg ( + ) ;Slc26a4Δ/Δ mice at P14 ( Fig . S2A ) , P16 ( Fig . 10A , C , D and Fig . S2C , G , K ) , P18 ( S2E ) , and P35 ( Fig . 10B ) . The absence of pendrin in Tg ( + ) ;Slc26a4Δ/Δ mice was observed with two different anti-pendrin antibodies ( Pds #1 and Pds #2 ) . In contrast , pendrin expression was found in transitional cells of the utricle , saccule and ampullae of controls that consisted of Tg ( − ) ;Slc26a4Δ/+ or Tg ( + ) ;Slc26a4Δ/+ mice . The expression patterns in Tg ( − ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4Δ/+ mice with both antibodies were similar to the pattern previously described in Slc26a4Δ/+ mice [11] , [12] . Vestibular labyrinths were isolated by microdissection from Tg ( − ) ;Slc26a4Δ/Δ , Tg ( + ) ;Slc26a4Δ/Δ and Tg ( + ) ;Slc26a4Δ/+ mice and the roof of the utricle was removed to permit an unobstructed view onto the utricular macula ( Fig . 11A–C ) . Glittering otoconia were observed in Tg ( + ) ;Slc26a4Δ/Δ and Tg ( + ) ;Slc26a4Δ/+ mice and giant otoconia in Tg ( − ) ;Slc26a4Δ/Δ . Otoconia were transferred into glass-bottom dishes and inspected by laser-scanning microscopy using a 405 nm laser . Giant otoconia from Tg ( − ) ;Slc26a4Δ/Δ mice were ∼10-fold larger than normal otoconia ( Fig . 11D ) . The shape of the giant otoconia resembled the shape previously observed in Slc26a4Δ/Δ mice [12] . Otoconia in Tg ( + ) ;Slc26a4Δ/Δ and Tg ( + ) ;Slc26a4Δ/+ mice were similar consisting in both genotypes of larger ( ∼20 µm , Figs . 11E . 1 and F . 1 ) and smaller ( ∼10 µm; Figs . 11E . 2 and F . 2 ) otoconia , some of which revealed a concentric structure ( Figs . 11E . 3 and F . 3 ) . These data suggest that the introduction of the transgene rescued normal otoconia formation . Balance tests were performed in Tg ( + ) ;Slc26a4Δ/Δ mice and Tg ( + ) ;Slc26a4Δ/+ mice as well as in Tg ( − ) ;Slc26a4Δ/+ and Tg ( − ) ;Slc26a4Δ/Δ mice ( Fig . 12 ) . Tg ( − ) ;Slc26a4Δ/Δ mice failed the test , which confirmed the vestibular phenotype previously described in Slc26a4Δ/Δ mice [14] . There was no apparent difference in the performance of Tg ( + ) ;Slc26a4Δ/Δ , Tg ( + ) ;Slc26a4Δ/+ and Tg ( − ) ;Slc26a4Δ/+ mice . These data demonstrate that the introduction of the transgene rescued normal gross motor vestibular function .
In this study we generated a mouse that expresses human pendrin in the endolymphatic sac but lacks detectable pendrin protein expression in the cochlea or in the vestibular labyrinth . The most biologically interesting and clinically relevant observation of this study is that this mouse develops normal hearing and balance . Our findings support the hypothesis that pendrin expression in the endolymphatic sac is chiefly responsible for the development of normal endolymph volume , that lack of pendrin in the endolymphatic sac is mainly responsible for the development of the membranous labyrinth enlargement in Slc26a4Δ/Δ mice , and that the complex inner ear pathology found in Slc26a4Δ/Δ mice is largely a consequence of the enlargement during embryonic development . This hypothesis was based on the studies in Slc26a4Δ/Δ mice that revealed that the enlargement is a key event on the path toward organ failure resulting in deafness and vestibular dysfunction [12] , [13] , [18] and on studies in Foxi1−/− mice that lack pendrin expression in the endolymphatic sac , develop an enlargement of the inner ear , but express pendrin in the cochlea and the vestibular labyrinth [19] . To test our hypothesis , we generated a transgenic mouse , Tg ( B1-hPDS ) , which expresses human SLC26A4 ( previously named PDS ) controlled by the promoter of ATP1V1B1 [15] . The ability of the promoter of ATP1V1B1 to control gene expression had previously been evaluated in a transgenic mouse that expresses eGFP controlled by the promoter of ATP6V1B1 [16] , [20] . Expression of eGFP had been found in this mouse in intercalated cells of the renal collecting duct , and in narrow and clear cells of the epididymal epithelium of adult mice [16] . We found expression of eGFP in the embryonic kidney and in mitochondria-rich cells of the endolymphatic sac but not in the cochlea or the vestibular labyrinth ( Fig . 3 ) . Expression in mitochondria-rich cells of the endolymphatic sac was expected since these cells are members of the FORE family ( forkhead related ) of cells . FORE cells express FOXI1 , which drives the expression of Atp6v1b1 and Slc26a4 [21] , [22] , [23] . Consistently , mitochondria-rich cells of the endolymphatic sac express the mRNAs Atp6v1b1 and Slc26a4 [24] , [25] and the corresponding proteins , the B1-subunit of the vH+ATPase and pendrin [11] , [23] . The onset of expression of Atp6v1b1 in the endolymphatic sac is at E11 . 5 , which is similar to the onset of pendrin [13] , [26] . Thus , it was likely that the transgene Tg ( B1-hPDS ) would drive a timely expression of pendrin in mitochondria-rich cells of the endolymphatic sac . Although FOXI1 drives the expression of Atp6v1b1 and Slc26a4 in FORE cells such as the mitochondria-rich cells of the endolymphatic sac , the expression of Atp6v1b1 and Slc26a4 is not limited to FORE cells . Indeed , Slc26a4 is expressed in the inner ear in spiral prominence and outer sulcus epithelial cells as well as in spindle-shaped cells of the cochlea and in transitional cells of the vestibular labyrinth , none of which are FORE cells [11] , [12] . Further , Atp6v1b1 expression has been found in the spiral limbus of the cochlea , which does not contain FORE cells [25] , [26] . The expression of Atp6v1b1 in the cochlea provided the possibility that the transgene would drive an ectopic expression of pendrin in the spiral limbus . Our studies of eGFP expression ( Fig . 3 ) and of pendrin expression by Western blotting ( Fig . 4 ) and immunocytochemistry of whole-mounted specimens and sections using two different anti-pendrin antibodies ( Pds #1 and Pds #2 , Fig . 7 and S1 ) revealed no detectable expression in the cochlea or vestibular labyrinth of Tg ( + ) ;Slc26a4Δ/Δ mice . The observed absence of pendrin expression in the vestibular labyrinth of Tg ( + ) ;Slc26a4Δ/Δ mice ( Fig . 10 and S2 ) is consistent with the reported lack of Atp6v1b1 expression in the vestibular labyrinth based on detection by in situ hybridization [25] , [26] and by quantitative RT-PCR ( Fig . 2B and F ) , which is a more sensitive technique . The observation that human SLC26A4 mRNA but no pendrin nor eGFP protein was detected in the cochlea or the vestibular labyrinth suggests the presence of strong translational regulation [27] . Taken together , our data demonstrate that we have generated a mouse that expresses pendrin in the endolymphatic sac but not in the cochlea or the vestibular labyrinth , although we cannot completely rule out that low levels of pendrin protein expression escaped our detection . Such low pendrin expression is unlikely the reason for the restored endolymphatic volume , since pendrin expression in the cochlea and vestibular labyrinth of Foxi−/− mice , which lack pendrin expression in the endolymphatic sac , did not prevent endolymphatic enlargement [19] and since mice that express a mutant pendrin protein that supports anion exchange at a reduced rate are deaf , develop mega-otoconia and are balance impaired [28] . Moreover , hypomorphic mutant alleles of SLC26A4 show no difference in the resulting auditory phenotype from that of functional null alleles in patients with Pendred syndrome [29] , indicating that small amounts of pendrin activity are insufficient to rescue hearing in humans . Measurements of the endocochlear potential and pH revealed that the introduction of the transgene , which rescued normal endolymph volume , also rescued the loss of the normal endocochlear potential and the loss of the normal endolymphatic pH homeostasis ( Fig . 8 ) . It appears that the Cl−/HCO3− exchanger pendrin , which is normally expressed in the apical membranes of spiral prominence and outer sulcus epithelial cells , is not the sole mechanism responsible for the alkaline pH of endolymph in a normally developed cochlea . A similar conclusion can be drawn based on measurements in the doxycycline-inducible Slc26a4 mouse model where termination of pendrin expression at P6 led to the development of a nearly normal endocochlear potential and of a nearly normal alkaline pH [30] . We hypothesize that the epithelial barrier enclosing endolymph is permeable to H+ , OH− and HCO3− and that the pH of endolymph follows the endocochlear potential . Hearing and balance tests in Tg ( + ) ;Slc26a4Δ/Δ mice revealed normal sensory function ( Fig . 9 and 12 ) . The observation that hearing thresholds at 32 kHz had some variability in Tg ( + ) ;Slc26a4Δ/Δ mice and that some Tg ( + ) ;Slc26a4Δ/Δ mice developed progressive high-frequency hearing loss is most likely a function of the genetic background . Tg ( + ) ;Slc26a4Δ/Δ were generated in a F2 generation from Slc26a4Δ/Δ mice that were maintained isogenic in the 129S6 background and Tg ( + ) ;Slc26a4Δ/Δ mice that were recently generated in a mixed background of C57BL/6 and CBA . DNA from the three background strains , 129S6 , C57BL/6 and CBA , which differ in their hearing thresholds , are expected to comprise variable amounts of the genomes of individual mice . Hearing thresholds for 1 to 3 month-old 129S6 , C57BL/6 and CBA mice range between 20–35 dB-SPL at 8 kHz , 10–28 dB-SPL at 16 kHz and 20–50 dB-SPL at 32 Hz [18] , [31] , [32] , [33] , [34] , [35] . In general , 129S6 , C57BL/6 and CBA mice have similar thresholds at 8 kHz , whereas at 16 and 32 kHz CBA mice have lower thresholds than 129S6 and C57BL/6 mice . Thus , the greater variability in hearing thresholds that was observed at 32 kHz particularly in Tg ( + ) ;Slc26a4Δ/Δ may be due to a variability in the mixture of these background strains . Our observation that normal hearing developed in the absence of pendrin expression in the cochlea in combination with the published finding that normal hearing was maintained when pendrin expression was terminated after completed development [30] , could suggest that pendrin in the cochlea has no physiologic significance beyond the developmental phase . However , it is also conceivable that pendrin-mediated HCO3− secretion provides a buffer that stabilizes the pH in the lateral wall tissues as well as in endolymph , and that this buffering is important during stress situations associated with normal life . Pendrin expression may indeed be important for the maintenance of hearing into advanced age . In summary , we demonstrated that restoration of pendrin to the endolymphatic sac is sufficient to restore normal inner ear function . This implies that pendrin in the endolymphatic sac is more important for the development of normal hearing than pendrin expression in the cochlea and more important for the development of normal balance than pendrin expression in the vestibular labyrinth . This finding , in conjunction with our previous report that pendrin expression is required for embryonic development but not for the maintenance of hearing , opens the prospect that a spatially and temporally limited therapy will restore normal hearing in human patients carrying a variety of mutations of SLC26A4 .
All animal experiments and procedures at Kansas State University were performed according to protocols approved by the Animal Care and Use Committees at Kansas State University ( IACUC#: 2961 ) . All animal procedures at Sorbonne University Paris Cité were performed according to protocols approved by the ethics committee from University Pierre et Marie Curie , and were performed in accordance with the Guide for the Care and Use of Laboratory Animals ( NIH publication No . 93-23 , revised 1985 ) . Human SLC26A4 cDNA was ligated into a pBluescript vector that contained 6 . 9 kbp of the human ATP6V1B1 promoter [16] , [20] . An SV40 late region polyadenylation signal was cloned downstream of the SLC26A4 cDNA . The transgene Tg ( B1-hPDS ) included the 5′-flanking region of the ATP6V1B1 gene extending to but excluding the endogenous translational start codon , the human SLC26A4 cDNA , with its own translational start site , and the SV40 late region polyadenylation signal . The integrity of the transgene was confirmed by restriction digest and bidirectional sequencing of ligation sites . In preparation for injection , the transgene was linearized by SalI and NotI digestion , followed by gel purification using an electroelution method and then concentrated using ElutipD columns ( Whatman ) . The transgene was then further concentrated by ethanol precipitation and resuspended in low EDTA injection buffer ( 10 mM Tris with 0 . 1 mM EDTA ) . Tg ( B1-hPDS ) transgenic mice were created by the University of Utah transgenic mouse core facility using standard procedures [16] , [20] . Genotyping revealed that 63 pups were positive for transgene integration . One founder , which transmitted the transgene in a Mendelian fashion , was crossed with wild-type C57BL/6× CBA F1 mice to establish a colony . Three Tg ( B1-hPDS ) ;Slc26a4+/+ transgenic mice were shipped to Kansas State University in Manhattan , Kansas , USA . At Kansas State University , a colony of Tg ( B1-hPDS ) Tg/+;Slc26a4Δ/Δ mice was established . Colony management was supported by software ( Litter tracker , written in Microsoft Visual Basic and Excel 2010 by P . W . ) Tg ( B1-hPDS ) Tg/+;Slc26a4+/+ mice were crossed with Slc26a4Δ/Δ mice to generate Tg ( B1-hPDS ) Tg/+;Slc26a4Δ/+ mice . Matings of Tg ( B1-hPDS ) Tg/+;Slc26a4Δ/+ mice generated 28 Tg ( B1-hPDS ) Tg/+;Slc26a4+/+ , 58 Tg ( B1-hPDS ) Tg/+;Slc26a4Δ/+ and 39 Tg ( B1-hPDS ) Tg/+;Slc26a4Δ/Δ mice in a near Mendelian ratio of 1 ∶ 2 ∶ 1 with a 75% rate of transmission for the transgene ( based on 169 pups ) . Mice were genotyped for Slc26a4+ and Slc26a4Δ alleles by PCR using established primers [14] and for the transgene Tg ( B1-hPDS ) ( Transnetyx , Cordova , TN ) . Primers for the transgene were designed to amplify a 345 bp PCR-product spanning the hPDS cDNA and the SV40 polyadenylation signal sequence ( left primer: 5′-aga ggg tca agg ttc cat ttt ag-3′; right primer: 5′-caa acc aca act aga atg cag tg-3′ ) [15] . Time-pregnant dams were deeply anesthetized with 4% tri-bromo-ethanol ( 0 . 014 ml/g body weight , i . p . ) and embryos were harvested by laparotomy . Dams and embryos were sacrificed by decapitation . Gestational age was counted from the day when the vaginal plug was detected . This day was set to embryonic ( E ) day 0 . 5 . Gestational age , however , was verified by evaluating gross morphological features including limbs , digits and the appearance of the pinna and auditory meatus [36] , [37] . The age of mice was counted from the day of birth , which was set to postnatal ( P ) day 0 . Postnatal mice were deeply anesthetized with 4% tri-bromo-ethanol ( 0 . 014 ml/g body weight , i . p . ) and sacrificed by decapitation or cardiac perfusion with fixative . Quantitative RT-PCR was performed on total RNA [17] . Total RNA was isolated from tissues obtained by microdissection from Tg ( − ) ;Slc26a4Δ/+ and Tg ( + ) ;Slc26a4Δ/Δ mice and subjected to quantitative RT-PCR using gene-specific primers for 18S rRNA as well as for mRNA coding for the α-subunit of the mouse Na+/K+ ATPase Atp1a1 , the B1-subunit of the mouse vH+ATPase Atp6v1b1 , mouse pendrin Slc26a4 and human pendrin SLC26A4 , which was introduced via the transgene . Postnatal mice were genotyped by PCR prior to tissue collection . Embryonic Tg ( − ) ;Slc26a4Δ/+ mice were generated by mating Tg ( − ) ;Slc26a4Δ/Δ dams and Tg ( − ) ;Slc26a4+/+ sires , which yielded 100% of the desired genotype . Embryonic Tg ( + ) ;Slc26a4Δ/Δ mice were generated by mating Tg ( + ) ;Slc26a4Δ/Δ dams and sires , which yielded Tg ( + ) ;Slc26a4Δ/Δ and Tg ( − ) ;Slc26a4Δ/Δ mice in a ratio of 3 ∶ 1 . Since embryonic mice could not be genotyped prior to tissue collection , the desired Tg ( + ) ;Slc26a4Δ/Δ mice among Tg ( − ) ;Slc26a4Δ/Δ mice were initially identified by visual inspection of the size of the endolymphatic sac and the presence of ‘glittering’ otoconia , This phenotypic identification was subsequently confirmed by the presence of human pendrin SLC26A4 transgene by RT-PCR . Tissues were obtained by microdissection . Endolymphatic sacs ( 8–10 endolymphatic sacs from 4–5 animals per sample ) were obtained from mice at age E17 . 5 . Cochlear ducts ( 4 cochlear ducts from 2 animals per sample and 2 cochlear ducts from 1 animal per sample ) were obtained from mice at ages E17 . 5 and P2 , respectively . Vestibular labyrinths ( 6 vestibular labyrinths from 3 animals per sample ) were obtained from mice at age P8 . Total RNA was isolated from microdissected tissues ( RNeasy micro kit , Qiagen , Valencia , CA , USA ) , treated with DNAse ( RNeasy micro kit ) , combined with RNA storage solution ( Applied Biosystems/Ambion , Austin , TX ) , adjusted to a concentration of 10 ng/µl , and stored at −80°C . Quantity and quality of total RNA were evaluated by microfluidic electrophoresis ( BioAnalyzer , Agilent , Santa Clara , CA ) , by microliter absorption photometry ( Nanodrop , Wilmington , DE ) and by quantitative RT-PCR of 18S rRNA . RNA samples were accepted for quantitative RT-PCR only when they were free of contamination and excellent RNA quality . RNA quality was quantified by the RNA integrity number ( RIN ) on a scale from 0 ( worst ) to 10 ( best ) ( BioAnalyzer , Agilent ) . RIN numbers for total RNA isolated from E17 . 5 endolymphatic sac and cochlea were 8 . 2±0 . 3 ( n = 3 ) and 9 . 2±0 . 1 ( n = 10 ) . Chemicals were assembled with the assistance of an automatic pipetting station ( Biomek NXp , Beckman Coulter , Fullerton , CA ) with hardware modifications and software programming by P . W . Quantitative RT-PCR reactions were carried out in 96-well plates with each well containing ∼10 ng of total RNA , gene specific primers , and an enzyme mix containing reverse transcriptase and DNA polymerase ( iScript , BioRad , Hercules , CA ) in a total volume of 25 µl . Reverse transcription was performed for 10 min at 50°C and terminated by heating to 95°C for 5 min ( OneStepPlus , Applied Biosystems , Foster City , CA ) . PCR consisted of 40 cycles of 10 s melting at 95°C , 30 s annealing and elongation at 58°C , and 15 s hot-measurement at 78°C ( OneStepPlus , Applied Biosystems ) . Left and right primers ( exon , product size ) were for 18S 5′-gag gtt cga aga cga tca ga-3′ and 5′-tcg ctc cac caa cta aga ac-3′ ( 316 bp ) , for Atp1a1 5′-tgc ccg cct caa cat tcc-3′ ( exon 14 ) and 5′-gac aca tca gag cca aca atc c-3′ ( exon 16 , 291 bp ) , for Atp6v1b1 5′-tga ccc gaa act aca tca cc-3′ ( exon 1 ) and 5′-gcc aga gcc att gaa aat cc-3′ ( exon 5 , 305 bp ) , for mouse Slc26a4 5′-tct gat gga ggc aga gat ga-3′ ( exon 20 ) and 5′-ggc cag cct aac aga gac ag-3′ ( exon 21 , 430 bp ) , and for human SLC26A4 were 5′-tcc caa agt gcc aat cca ta-3′ and 5′-aca tca agt tct tct tcc gtc ag-3′ ( 360 bp ) . Primer pairs for Atp1a1 , Atp6v1b1 , mouse Slc26a4 spanned introns to prevent amplification of genomic DNA . Primer pairs for mouse Slc26a4 detected the Slc26a4+ allele as well as the Slc26a4Δ allele that is lacking exon 8 [14] . Left and right primers for mouse Slc26a4 differed by 7 and 10 nucleotides from the corresponding human sequence and left and right primers for human SLC26A4 differed in 4 and 6 nucleotides from the corresponding mouse sequence , thereby maximizing species-specific amplification . Since the human transgene did not contain introns , some reactions were carried out without reverse transcriptase to determine whether products of SLC26A4 originated from cDNA rather than from genomic DNA . These experiments revealed no evidence for significant amplification of genomic DNA . Amplification of a single product of the appropriate size was verified by microfluidic electrophoresis ( BioAnalyzer , Agilent ) . The number of template molecules ( cDNATemplate ) was estimated according towhere 6 . 02×1023 molecules/mol represents Avogadro's number , ProductThreshold is the weight of the PCR-product at threshold ( 0 . 49×10−9 g ) that was obtained from calibration experiments , ProductSize is the size of the product in base pairs ( bp ) , Weightbp is average weight of one bp ( 660 g/mol ) , Efficiency is the PCR-efficiency obtained from the slope of the log-linear phase of the growth curve [38] and Ct is the cycle at which the fluorescence of the product molecules reaches a common threshold chosen in the middle of the log-linear part of the growth curve . Bisected heads of embryos age E15 . 5 were fixed overnight in Bodian's fixative , contained ( vol/vol ) 75% ethanol , 5% acetic acid , and 5% formalin in water . Heads were then dehydrated overnight in 100% ethanol and cleared in methyl-salicylate [39] , [40] . The membranous labyrinth was injected via the lateral wall of the basal turn of the cochlea and via the endolymphatic sac with diluted paint ( Liquid Paper , Newell Rubbermaid , Atlanta , GA , 0 . 1–0 . 2% in methyl-salicylate ) using a fine glass-electrode , a manipulator ( NM-151 Narishige ) and a micrometer-driven oil-filled microinjector ( CellTram Vario , Eppendorf , Hamburg , Germany ) . For each genotype , at least three inner ears were injected . Whole mounts of fresh cochlear ducts , endolymphatic sacs and slices of kidney were prepared from E15 . 5 Tg ( B1-eGFP ) mice and visualized with a fluorescence microscope ( AxioScope , Carl Zeiss Göttingen ) . Mice were anesthetized with 4% tri-bromo-ethanol for in situ measurements of the endocochlear potential and pH with double-barreled microelectrodes . Measurements were made in the basal turn of the cochlea by a round-window approach through the basilar membrane of the first turn of the cochlea [18] , [42] . The surgical cavity was covered with liquid Sylgard 184 ( Dow Corning ) to limit the loss of tissue CO2 into ambient air . Double-barreled glass microelectrodes were pulled ( micropipette puller PD-5; Narishige ) from filament-containing glass tubing ( 1B100F-4; World Precision Instruments ) and baked at 180°C for 2 h to ensure dryness . One barrel was silanized by a 30 s exposure to 0 . 008 ml dimethyldichlorosilane ( 40136; Fluka ) . After silanization , microelectrodes were baked again at 180°C for 3 h and tips were broken to a final O . D . of ∼3 µm . The reference barrel was filled with 150 mM KCl and the ion-selective barrel was filled at the tip with liquid ion exchanger ( Hydrogen ionophore II - Cocktail A , 95297; Fluka ) and back-filled with buffer solution ( 500 mM KCl , 20 mM HEPES , pH 7 . 4 ) . Each barrel of the double-barreled microelectrode was connected via a Ag/AgCl2 electrode to an electrometer ( FD223 , World Precision Instruments ) . A flowing KCl electrode ( 1 M KCl in 0 . 2% agar ) was inserted under the skin of the animal to serve as ground electrode . Data were recorded in analog ( BD12E Flatbed recorder , Kipp & Zonen , Delft , The Netherlands ) and digital form ( DIGIDATA 1322A and AxoScope 10 , Axon Instruments , Union City , CA ) . pH electrodes were calibrated in situ at 37°C using three calibration solutions with different pH values . Calibration solutions contained ( in mM ) : pH 6: 130 NaCl , 20 MES; pH 7: 130 NaCl , 20 HEPES; and pH 8: 130 NaCl , 20 HEPES . pH-sensitive electrodes had a slope of 56 . 9 ± 0 . 3 mV/pH unit ( n = 11 ) . Mice were deeply anesthetized with a mixture of dexmedetomidine and ketamine ( 0 . 375 mg/Kg body weight dexmedetomidine and 56 mg/kg body weight ketamine; i . p . ) and placed on a thermal pad to maintain normal body temperature . The mastoid , vertex and ventral neck region of the animal were connected via sub-dermal platinum needle electrodes ( F-E2 , Astro Med , Rhode Island , RI ) and short ( 31 cm ) leads to the main channel , reference channel and ground of the preamplifier , respectively . Auditory brainstem recordings were performed in a custom constructed , electrically shielded and sound-attenuated chamber ( inner dimensions: 23 cm×23 cm×23 cm ) using a digital data acquisition system ( BioSig32 software , RA4LI Preamplifier , RP2 . 1 Enhanced Real Time Processor , PA5 Programmable Attenuator , ED1 Electrostatic Speaker Driver , Tucker-Davis Technologies , Alachua , FL ) . Tone burst stimuli were presented ( 21 per sec ) via a free field electrostatic speaker ( SigGen software , ES1 speaker , Tucker Davis ) . Acoustic stimuli were calibrated using a 1/4 inch condenser microphone ( SigCal IRP4 . 2 software , Tucker Davis , PS9200 microphone , Acoustical Interface , Belmont , CA ) placed at the location of the mouse head . Tone bursts ( 2 ms duration , 0 . 5 ms gate time; 8 , 16 and 32 kHz ) were presented with alternating phase ( 0 and 180° ) . Responses , recorded over 10 ms , were filtered ( 300 Hz high pass , 3000 Hz low pass and 60 Hz notch ) and 1000 recordings were averaged . Tone burst stimuli were presented at intensities varying between 90 and 0 dB SPL in 5 dB intervals . Auditory thresholds were obtained by a visual comparison of wave forms . After the procedure , mice were rapidly recovered from anesthesia with atipamizole ( 1 . 875 mg/kg body weight; i . p . ) . Balance testing consisted of determining the time that mice could balance on a rotating 1″ rod with rotations ramping up from 4 to 40 rpm in 60 s ( RotaRod , IITC Life Science , Woodland Hills , CA ) . Test chambers were cushioned with bubble-foil to provide a soft landing for mice falling off the rod .
|
Mutations of SLC26A4 are the most common cause for hearing loss associated with a swelling of the inner ear . This human disease is largely recapitulated in a mutant mouse model . Mutant mice lack Slc26a4 expression and their inner ears swell during embryonic development , which leads to failure of the cochlea and the vestibular organs resulting in deafness and loss of balance . SLC26A4 is normally found in the cochlea and vestibular organs of the inner ear as well as in the endolymphatic sac , which is a non-sensory part of the inner ear . The multitude of sites where SLC26A4 is located made the goal to restore function through restoration look futile , unless some sites were more important than others . Here , we generated a new mutant mouse that expresses SLC26A4 in the endolymphatic sac but not in the cochlea or the vestibular organs of the inner ear . Fantastically , this mouse did not develop the detrimental swelling of the inner ear and even more exciting , the mouse developed normal hearing and balance . Our study provides the proof-of-concept that a therapy aimed at repairing the endolymphatic sac during embryonic development is sufficient to restore a life-time of normal hearing and balance .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"developmental",
"biology",
"otology",
"genetic",
"mutation",
"gene",
"expression",
"genetics",
"molecular",
"genetics",
"hearing",
"disorders",
"biology",
"otorhinolaryngology",
"morphogenesis",
"genetics",
"of",
"disease",
"gene",
"function"
] |
2013
|
SLC26A4 Targeted to the Endolymphatic Sac Rescues Hearing and Balance in Slc26a4 Mutant Mice
|
Recently , thanks to the increasing throughput of new technologies , we have begun to explore the full extent of alternative pre–mRNA splicing ( AS ) in the human transcriptome . This is unveiling a vast layer of complexity in isoform-level expression differences between individuals . We used previously published splicing sensitive microarray data from lymphoblastoid cell lines to conduct an in-depth analysis on splicing efficiency of known and predicted exons . By combining publicly available AS annotation with a novel algorithm designed to search for AS , we show that many real AS events can be detected within the usually unexploited , speculative majority of the array and at significance levels much below standard multiple-testing thresholds , demonstrating that the extent of cis-regulated differential splicing between individuals is potentially far greater than previously reported . Specifically , many genes show subtle but significant genetically controlled differences in splice-site usage . PCR validation shows that 42 out of 58 ( 72% ) candidate gene regions undergo detectable AS , amounting to the largest scale validation of isoform eQTLs to date . Targeted sequencing revealed a likely causative SNP in most validated cases . In all 17 incidences where a SNP affected a splice-site region , in silico splice-site strength modeling correctly predicted the direction of the micro-array and PCR results . In 13 other cases , we identified likely causative SNPs disrupting predicted splicing enhancers . Using Fst and REHH analysis , we uncovered significant evidence that 2 putative causative SNPs have undergone recent positive selection . We verified the effect of five SNPs using in vivo minigene assays . This study shows that splicing differences between individuals , including quantitative differences in isoform ratios , are frequent in human populations and that causative SNPs can be identified using in silico predictions . Several cases affected disease-relevant genes and it is likely some of these differences are involved in phenotypic diversity and susceptibility to complex diseases .
Alternative splicing ( AS ) allows for multiple mRNA isoforms to be transcribed from a single gene locus , potentially creating much greater protein diversity from our roughly 25 thousand human genes [1] . AS is very common in higher order organisms and , especially through the lens of newer , high throughput technologies , such as oligonucleotide arrays and transcriptome sequencing , we are finally realizing the true extent and importance of AS . New studies based on deep sequencing and micro-arrays with exon junction probes estimate the proportion of genes undergoing AS in humans between 74% and 94% [2]–[5] Hence , it is crucial to understand the functions and the regulation of AS if we are to arrive at a real grasp of regulation of gene expression and gene networks . Although most isoform variation is thought to occur between tissues , many differences exist among healthy individuals in a population [6]–[8] . It is likely these differences are of genetic origin and contribute to phenotypic diversity and disease susceptibility . Many Mendelian disorders , such as cystic fibrosis [9] , have been explained by splicing errors caused by genetic mutations [10] . This shows the importance of finding more genetically driven isoform variations to understand the genetic causes of complex diseases . Until recently , AS differences were not detectable with commercially available micro-array platforms . Due to low probe densities , those platforms only aimed at measuring gene-level expression and targeted mainly the 3′ untranslated region ( UTR ) of genes . The Affymetrix Exon Array , with its nearly 5 . 5 million exon-targeted probes , is one of the recent genomic tools available for profiling of splicing or transcript initiation/termination differences between human tissues or between individuals . In the prelude to this work , Kwan et al . [6] showed using Exon-Array expression data from lymphoblast cell-lines of HapMap individuals that cis-acting polymorphisms are associated with many gene-level expression differences and isoform ratio differences between individuals of the HapMap CEPH population [11] . While the initial analysis detected numerous robust differential splicing events that were genetically controlled , the study did not attempt to identify the actual causal polymorphisms and it was not clear whether it had sufficient statistical power and signal to noise ratio to detect more subtle genetic influences on exon inclusion levels . Given the high validation rate achieved for AS in the first study , we expected to find many more inter-individual splicing differences deeper , within the statistically less significant candidates and also within the speculative content of the microarray , which targets predicted or rarely expressed exonic regions . Moreover , we aimed to determine whether it was possible to identify the causative polymorphisms – as opposed to extended regulatory haplotypes – responsible for such changes . Thus , in order to increase our detection threshold , in the analysis presented here we made use of publicly available , prior AS information . Using a customized heuristic optimized for the detection of AS events and some visual curation , we selected a large sample of candidate genetically regulated AS events . Subsequently , we validated these events using RT-PCR , and in order to identify likely causative SNPs , we re-sequenced the genomic DNA around the alternatively spliced regions . Finally , we used in silico predictions and , in selected cases , minigene assays , to verify the causative nature of the detected polymorphisms . This new analysis on the data demonstrates that the Exon Array can detect much more subtle splicing differences than initially suggested and that the speculative , non-core probesets , which constitute the majority of the array , contain useful data for AS discovery . This study constitutes the most in-depth analysis of cis-regulatory heritable splicing differences to date , having validated quantitatively the greatest number of cases and found a likely causative SNP is most cases . Our results provide new insights in the genetic regulation of splicing , demonstrate that subtle differences in alternative splicing between humans are more frequent than initially detected , and that the causative polymorphisms can be identified and validated in reporter mini-gene systems .
To optimize our chances of finding true AS events , we only considered candidate probesets whose target genomic coordinates overlapped an AS event catalogued in at least one database , including KnownAlt [12] , ASAP-II [13] and additional events inferred directly from EST/mRNA genome annotations ( see Methods ) . We ranked candidates statistically according to both the regression p-value and a custom-designed measure representing the unexpectedness of the probeset's fold-change in the context of the other probesets in the gene ( see Methods ) . Further , by visually inspecting probeset fold-changes and regression P-values of the top candidate genes in the UCSC Genome Browser , we selected 68 new potential AS events , along with 4 events from previous analyses [6] , for further validation and characterization . For the purpose of this analysis , we chose events that would be easily amplifiable via PCR , therefore excluding large intron retentions or alternative transcript initiation/termination events . The UCSC Genome Browser Tracks of selected candidates are available in Table S1 . The red “AS-marker” track indicates the position of the affected exon . In order to detect isoform ratio differences quantitatively as well as qualitatively , we validated our candidate events using semi-quantitative RT-PCR . Instead of performing the validations on all 57 individuals , we chose a strategic sub-sample of 10 individuals from the HapMap CEPH population , which was the smallest possible sample for which at least one individual was polymorphic at every SNP associated with an AS event . The primers were targeted to the two exons flanking each alternatively spliced region . Product abundance was quantified as described in the Methods . Figure 1A and 1B show examples of electrophoresis readings from which the isoform ratios were estimated . Out of the 58 candidates which produced interpretable results , 22 showed a significant association with the SNP ( p<0 . 05 , Figure 1B shows an example ) , confirming that the AS event is under the control of a cis-regulatory mutation , 10 showed a visible trend in the expected direction , 7 showed clear differences in isoform ratios between individuals but with no obvious trend linked to the SNP genotype , and 3 showed visible isoforms with no detectable differences in ratios . The remaining 16 showed no evidence of AS . It should be noted that the small sample size ( 10 individuals ) and limited representation of different genotypes limits the power of this validation approach . Thus , while observing a single PCR product in all samples can be considered as a reliable indication of a false positive result , observing two alternatively spliced products of the expected sizes , even without achieving statistical significance , is evidence supporting the initial microarray finding . Thus depending on the stringency of the validation criteria – detecting a statistically significant association within the PCR data , versus observing two alternatively spliced products - our validation rate is between 38% and 72% . Table 1 shows all validated AS events organized by the strength of the validation evidence . For all the candidates selected in this study and a few additional validated candidates from the previous analysis [6] , we sequenced a region of 600–800 bps around the putative AS events in 2 individuals predicted to preferentially express one isoform and 2 individuals expressing the other . The selection of individuals was based on both the genotype of the associated SNP and the micro-array expression scores for the probeset of interest . Thus , even if the associated SNP was not perfectly linked to the causative SNP , selecting for extreme expression phenotypes increased the chances that the 4 individuals would differ at the causative polymorphism . Analysis of the sequencing results revealed 86 polymorphisms in 60 confidently sequenced regions , 76 of which were already catalogued in dbSNP129 [14] , and 10 which were novel . Out of the 34 validated AS events which were successfully sequenced ( including 4 from the previous analysis ) , all showed at least one SNP within the sequenced interval , as did 6 out of the 7 sequenced regions which were negatively validated . For each gene in which one or more SNPs were identified , Table S2 links to custom UCSC Genome Browser tracks which indicate the position of SNPs and the associated fold-change in sequenced individuals . For 24 of the genes , we could identify a SNP within 20 bases of a relevant splice-site . Using the MaxEntScan algorithm [15] , which calculates the theoretical strength of splice-sites based on maximum entropy , we could verify whether the expected effects of many SNPs adequately explained the difference in micro-array expression between the sequenced individuals . In all 17 cases where differences in splice site strength could be calculated , the micro-array results and the maximum entropy scores of the polymorphic splice-site sequences agreed on the direction of the effect ( See Table 2 ) . Figure 2 shows the types of AS events and the relative position of the affected splice-site , which is crucial to understanding the direction of the expression changes . For an additional 16 cases where a SNP was present within the affected exon , we used the ESE Finder 3 . 0 online tool [16] to predict ESEs affected by the SNP and assign them a matrix-based affinity score ( see Table 3 ) . In 13 out of the 16 exons , the predicted change in the number and/or affinity of ESE motifs was concordant with the probeset expression change . Although , in some of these cases , identifying the affected splice-site is not obvious , we are still able in all cases to infer the direction of the predicted probeset expression change , as shown in Figure 2 . The 3 for which the predicted effect was in disagreement were the 3 cases with the smallest expression fold-change . This result is encouraging , particularly since it only concerns the binding preferences of 4 splicing factors , out of possibly dozens , and the less than perfect agreement between predictions and results likely reflects the relative lack of detailed understanding of ESE motifs as compared to splice-site consensus sequences . Out of the 4 validated and sequenced AS regions remaining , 1 had a SNP inside the retained intron ( HNRPH1 ) and 3 had one or more intronic SNPs around the cassette exon ( IL6 , WDR67 and SIDT1 ) . It is likely that intronic SNPs play a role in determining isoform levels through their effect on intronic splicing enhancers ( ISEs ) or silencers ( ISSs ) . We used another online software tool , SpliceAid [17] , to detect ISEs or ISSs but only one of the 3 cases , showed qualitative agreement with the expression data ( data not shown ) , indicating that we are either looking at the wrong candidate SNPs or that our understanding of ISEs/ISSs is not yet detailed enough to predict the effect of all these polymorphisms . Since we only sequenced approximately 200 base pairs from each exons , and introns generally span thousands of bases , causative SNPs are likely to be found further away in the intron . Intronic SNPs were found in many cases in conjunction with exonic SNPs , but there were often multiple intronic SNPs . Since we only looked for qualitative agreement between in silico predictions and the observed splicing differences , looking at many SNPs per gene would surely have caused more chance correlations and we would have been unable to rank the effects of SNPs affecting different splicing factors . For this reason , we prioritized exonic or splice-site bordering SNPs , for which there was almost always a single candidate . In order to assess whether the candidate SNPs truly cause the splicing differences in vivo , for 6 genes we sub-cloned the exon and surrounding intronic sequence from individuals of different genotypes into a minigene expression vector [18] . Placing the exon of interest between two constitutive exons within the reporter construct allows determining the effect of the SNP on candidate exon inclusion levels . We used sequencing to verify that the sub-cloned construct differed only at the SNP positions we had previously predicted to be responsible for the differential splicing event . In 5 out of the 6 cases , the putative causative SNP was the only SNP present . MMAB however contains 3 closely neighboring SNPs which are perfectly linked . We transiently transfected the constructs into HeLa cells and assessed the presence of different isoforms using RT-PCR . Figure 3 shows the electrophoresis band migrations along with the SNP genotypes . For CAST , ERAP2 , and PARP2 , for which the candidate SNP is very close to the splice-site , we demonstrate that the predicted SNP causes a complete switch in 5′ splice-site usage ( PARP2 , ERAP2 , Figure 3A and Figure 2B ) or a complete skipping of the exon ( CAST , Figure 3C ) . In the 3 other cases , ATP5SL , MMAB and AMACR , for which the candidate SNPs were found in the exon and were predicted to disrupt ESEs , the assay shows a visible but subtle change in isoform ratios ( Figure 3D ) in the expected direction . We used the Agilent 1000 DNA chip to quantify the results for these three cases because their gel bands are less convincing than the first three cases . We included the capillary electrophoresis readings as Figure S1 . The quantification of the isoforms from the peaks show that the smallest difference between individuals of different genotypes , the first and last columns for the AMACR gene in Figure 3 , consists of a 1 . 5 fold difference in isoform ratios , confirming that all the changes in isoform ratios were measurable and in the expected direction . Why significant variation exists between individuals with the same genotype is hard to say , especially considering that the plasmid inserts were confirmed to have the same sequence . The differences must come either from changes incurred during transfection ( a single transfection assay was performed for each plasmid ) or from biological or technical noise . Except for MMAB , for which any one of the 3 SNPs or even a combination of all 3 could be causative , these results demonstrate the causative nature of our candidate SNPs and confirm that both systematic splicing differences as well as more subtle , quantitative , differences exist between individuals and the extent of the change is reflective of the position of the causative SNP and of in silico predictions of its effect . In order to gain insight into the persistence within human populations of SNPs with such drastic effects on the structure of the expressed transcripts , we wanted to assess whether some of the causative SNPs showed signs of recent positive selection . We performed two tests . Using the program fdist2 [19] , we calculated the fixation index ( Fst ) of all 41 re-sequenced HapMap SNPs based on the frequencies in the 4 available populations ( see Methods ) , and compared the results to a simulated dataset . The average Fst was significantly greater than expected ( p = 0 . 046 ) by a small margin . One SNP had a Fst above the 99 . 999th percentile relative to the simulation dataset , rs6580942 ( A/C ) , the best candidate SNP in the ESPL1 gene . This is very significant , even in a sample of 41 ( p = 4 . 1E-3 ) . The C allele has a frequency of 0% in the African ( YRI ) and Asian ( CHB+JPT ) populations compared to 33% in the Caucasian population ( CEU ) . We then calculated the highest relative extended haplotype homozygosity ( REHH ) in the CEU population considering various size blocks below 500 kb centered on the same 41 SNPs ( see Methods for details ) . We performed the same analysis on 300 randomly chosen SNPs in genes as a control . There was no significant difference between the average highest REHH of our sequenced SNPs and the control set . No SNP showed a REHH score above the 99th percentile of the control set . However , we noticed that rs10941112 ( A/G ) , the best candidate SNP from the AMACR gene , had both a Fst above the 97th percentile , with the A genotype varying from 0% in the African population to 37% and 62% in the Asian and Caucasian populations respectively , and a REHH above the 97% percentile , with the A allele showing 122 . 5 fold greater homozygosity for a block of 391 kB around the SNP than the G allele . The combined extreme result in both tests is highly unlikely ( p = 8 . 0E-4 ) and is indicative of positive selection . These two analyses have brought to light strong evidence suggesting that the SNPs rs6580942 and rs10941112 , in the ESPL1 and AMACR genes , respectively , have undergone recent positive selection . Other than these 2 SNPs , there is no strong evidence of positive selection , indicating that many of our re-sequenced HapMap SNPs may be selectively neutral .
Previous micro-array studies have often attempted to estimate the real amount of AS or gene-level differences simply from counting the number of cases which surpass a multiple-testing corrected significance threshold [6] , [8] , placing complete faith in the results of the micro-array as well as the normalization , summarization and whichever AS detection algorithm was used . The first problem with such an approach is that many sources of noise can cause false positives , like SNPs within probes , cross-hybridizations or technical noise , as well as other distortions such as probesets responding unevenly to a gene-level change . The second weakness of such estimations , as our results demonstrate , is that many real AS events lie far below standard significance thresholds . The highest P-value for a validated AS candidate in this study was 7 . 3E-4 as opposed to 4 . 2E-9 in the previous study on the same dataset . This means that approximately 15 times more probesets could be considered potential AS candidates . Of course , we do not suggest that we found 15 times more AS than the previous study but rather , by integrating EST evidence and utilizing a more sophisticated AS detection algorithm , we can show that the speculative content of the array , which comprises about 80% of the array , as well as the less significant measurable differences on the array contain valuable information which have thus far , been mostly overlooked . Our results show that Exon-Array data by itself may be too noisy to produce reliable estimates of AS at the genome-wide scale . All we can conclude is that genetically controlled splicing differences exist between individuals and are probably more common than was previously estimated . A more definitive answer on the real extent of individual-specific AS may come from high-throughput sequencing , which can avoid probe target bias and detect AS differences qualitatively rather than through statistical inference . We have shown that many splicing differences between healthy individuals can be identified using the Exon Array platform . We expect that many more , perhaps approaching the complete map of the splicing eQTL ( expression quantitative trait loci ) landscape , will be catalogued soon using more sensitive methods , such as deep mRNA sequencing . Most of these splicing differences which we can detect are controlled by polymorphisms in cis-regulatory regions or in the vicinity of an implicated splice-site . These differences between individuals could contribute to phenotypic variation and could either be neutral in their effects , or confer differential susceptibility to complex diseases . A few of our validated events occur in genes which have already been associated with diseases . BCKDHA is related to maple syrup urine disease , type 1a [20] , a rare inherited metabolic disorder which , without a highly controlled diet and close monitoring of blood chemistry , causes progressive neurological damage which can cause vomiting , eating difficulties , irregular breathing , coma or death . Deficiency of the gene alpha-methylacyl-coa racemase ( AMACR ) is a rare disorder of the fatty acid metabolism which is characterized by neuronal and liver abnormalities [21] and the gene is considered a useful biomarker for various types of cancer [22] , making it quite interesting that this gene contained the SNP with the most evidence of positive selection . Interleukin 6 ( IL6 ) is an important mediator of fever [23] and the gene has been associated with osteoporosis [24] and Kaposi's sarcoma [25] . MMAB is related to vitamin B12 responsive methylmalonic aciduria [26] , the inability to synthesize adenosylcobalamin , a vitamin B12 derivative , and whose symptoms include metabolic acidosis and retarded development . A SNP in MMAB , which is in linkage disequilibrium with our causative splicing SNP , has recently been associated with HDL cholesterol levels [27] . Surprisingly , all of the above AS events are within protein-coding regions of the genes , making it very likely that these heritable differences contribute to individuals' predisposition to disorders similar to those caused by inactivation of those genes or to other , more complex , diseases . Complex diseases such as diabetes , cancer or schizophrenia are expected to be influenced by polymorphisms in a large number of genes , which may interact in multifarious ways . Many of the polymorphisms we identified in this study induce , through AS , potentially much more dramatic changes to the protein sequence than do non-synonymous coding SNPs . We have shown that common SNPs can influence alternative splicing across individuals even in relatively important genes , indicating that the genotyping of such SNPs will likely play a very significant role in predicting the occurrence of complex diseases in the future . Hundreds of genome-wide association studies ( GWAS ) carried out to date have generally failed to identify causative protein-coding disease variants [28] . Many of the underlying causes may be due to subtler , regulatory genetic influences [29] . Thus , a lot of interest and resources have been allocated to identifying eQTLs , genes whose expression levels are affected by regulatory SNPs . Although identifying eQTLs has been quite successful [6] , [30]–[33] , there has been considerably less accomplishments in teasing apart regulatory haplotypes and pinpointing the SNP actually responsible for the regulatory difference . Our group's earlier work demonstrated the existence of common isoform eQTLs; i . e . genes under genetic control resulting in differential expression of transcript isoforms including: alternative splicing , alternative polyadenylation , and alternative transcript initiation . Here , we postulate that in many of those cases , it should be possible to narrow down the regulatory region to the vicinity of the alternative event ( e . g . cassette exon ) , and subsequently identify the causative polymorphism with high confidence . In some of the cases , as in the ERAP2 gene , the common splicing polymorphism introduces a premature stop codon , most likely resulting in nonsense-mediated decay of the alternative product and a drastic reduction in the overall transcript levels , suggesting that splicing and RNA processing variations may be underlying some of the common expression QTLs . This knowledge will be essential for future functional studies and perhaps for future applications such as genetic therapy . In 2004 Nembaware et al . used public EST data to show the existence of allele-specific transcript isoforms in human [34] . In 2007 , Hull et al . showed that it is possible to identify such events in lymphoblastoid cell lines [7] . They selected 70 alternative splicing events , and showed that 6 of them were consistently associated with a specific genotype . They also used in vivo assays demonstrating causative nature of 2 candidate SNPs , suggesting that a substantial number of alternative splicing events may be controlled by genetic polymorphisms . These studies were followed by genome-wide microarray-based approaches [6] , [8] , [35] , which confirmed and further expanded our knowledge of genetic control of isoform variation . Earlier this year , Zhang et al . published an article [8] where they used the Exon-Array on lymphoblast cell lines to look for genetic variants which account for AS differences between populations . They claim , based on multiple-testing-corrected statistics , that they discovered 397 such differences between the Caucasian ( CEPH ) and African ( YRI ) populations . Recently , other groups have approached the problem from the purely genetics angle: knowing the polymorphisms that are present , they attempted to predict their effect computationally and identify exons that are differentially spliced across individuals . This approach has so far met with limited success . Elsharawy at al . [36] have obtained an extremely low validation rate of their in silico predictions , ranging from 0% for ESE predictions to 9% for SNPs in splice-sites , demonstrating our far from complete understanding of the effects of cis-regulatory sequences on splicing . Thus , in the present study , we take further steps towards optimal integrated use of the existing data – gene structure annotation , splicing-sensitive microarray data , SNP databases and targeted sequencing - to detect splicing eQTLs and their genetic determinants . First , taking advantage of the high level of coverage of the current sequence-based annotation of AS events , we concentrate only on events that have been previously reported . This approach is highly justified by the observation from the previous analysis [6] that less than 10% of the detected and validated AS events were novel ( unannotated ) , suggesting that the current EST coverage of the transcriptome is nearly complete . Secondly , we use a much improved algorithm to detect AS events in exon array data . Finally , we show that among the events that are detected using the above criteria and further validated in the lab , a majority contain SNPs that have highly suggestive in silico evidence of causation . We can show in a post hoc analysis that knowing the sequence variation information before-hand could have significantly improved the specificity of our search for AS . 27 out of 34 ( 79% ) validated and sequenced alternatively spliced regions contained a SNP for which in silico evidence appropriately explained the change , compared to 2 out of 7 ( 29% ) for the sequenced regions which were negatively validated ( data not shown ) . Given the current low resolution of HapMap SNPs , making use of in silico predictions of SNPs' effects at the genome-wide level would only be applicable to a fraction of Exon Array probesets . Less than half of our likely causative SNPs ( in splice-sites or exons ) were HapMap SNPs . The experience of Elsharawy et al . showed that the specificity of the purely computational approach is quite low , given the current level of understanding of AS regulation . However , once the resolution of SNPs and their genotypes increases significantly , which is already the case for four CEPH HapMap individuals which were recently fully sequenced ( http://www . 1000genomes . org/ ) , it should become feasible to merge computational predictions with biological expression data to improve the power to detect cis-acting polymorphisms involved in splicing . In turn , the identified polymorphisms and their effects will help to further enhance our understanding of splice-sites and cis-regulatory sequences . Out of the two splice-sites , the 5′ intronic splice-site ( the donor site ) appears to be the dominant identifiable target of these regulatory polymorphisms , with 14 predicted causative SNPs , as opposed to 3 for the 3′ splice-site ( the acceptor site ) . Given that we considered 23 bases for the 3′ splice-site compared to 9 , for the 5′ splice-site , it seems unlikely the result of pure chance , suggesting either that there may be greater purifying selective pressure acting on the region around the 3′ splice-site , or that , due to the greater degeneracy of the 3′ sequence [15] , SNPs affecting it influence splicing strength too subtly to be detected by the micro-array platform . In the case of cassette exons , the fact that the 5′ splice-site sequence , as well as ESE sequences , influence the use of the upstream splice-site defining the exon start , demonstrates the fact that in multi-intronic genes in vertabrates , the “exon definition” step , whereby splice-sites are paired across the exon , takes place before the assembly of the mature spliceosome and splicing of the intron can occur [37] , [38] , as opposed to species like yeast , whereby the intron definition occurs independently for each intron and a mutation of the 5′ splice-site would cause retention of the downstream intron rather than exon skipping [39] . The fact that the vast majority of putative causative SNPs affected either the 5′ splice-site or ESE sequences shows that the exon definition plays a central role in generating these splicing variations across individuals . Expression QTL analysis has garnered considerable interest in recent years and is increasingly being used in conjunction with whole genome association studies to narrow down the list of genetic variants putatively responsible for complex genetic disorders [28] . Here , we focus on a specific type of eQTL , alternative splicing variation , and extend the results of prior studies by validating the greatest number of these differences and showing that such variation may be more common than previously estimated , and that its effects can be quantitatively very subtle . Furthermore , this work demonstrates that we have the ability to identify the specific causative genetic variants responsible for isoform eQTL differences among individuals . It also underscores the value of data integration in order to obtain improved true positive rates in large scale analyses of splicing . With the upcoming release of the 1000 genomes data [40] and the growing use of high-throughput sequencing for transcriptome analysis , we will be able to broaden our understanding of the complex intricacies of the “splicing code” and use this knowledge to confidently identify splicing regulatory SNPs .
In order to select alternative splicing candidates within the vast set of significant associations between SNPs and probeset expression , we used publicly available knowledge of AS events to prioritize the list . We only considered Exon-Array probesets whose target coordinates overlapped an EST-supported AS event . We used a cut-off of a single base , reasoning that a single base mismatch could significantly affect the probe binding efficiency [41] . We downloaded the list of putative AS events in Human based on UCSC Known genes and human AS events from ASAP-II . We also retrieved from the UCSC Genome Browser website the full tables describing the human genome annotations based on Blated [42] spliced EST and mRNA sequences to gather some additional AS evidence . We applied strict criteria when inferring potential AS events from EST/mRNA data . To avoid confusing intron retention events with incomplete splicing or transcript length changes with incomplete mRNA sequences , we only selected cassette exons or alternative splice site usages . The latter two types of AS are also most confidently validated via PCR since they generally introduce short changes in the mRNA sequence . We had to define a reasonable gene structure from the Blat results , which , especially for EST annotations , include many short gaps which most probably originated from sequencing errors rather than genuine splicing . We considered gaps greater than 30 bps as introns , ignored gaps smaller or equal to 3 bases , and filtered out annotations containing gaps of any size in between . An exon was defined as a continuous alignment of 15 bases or longer , surrounded in the contig by an intron and a part of an exon on each side . Alternative splice-site usages had to be at least 6 bases long and their length a multiple of three if they fell within a UCSC gene's coding region . We insisted that EST's or mRNA's supporting an event should contain sequence from the next exon on each side of the event with the flanking introns spliced . Since we considered EST data to be less reliable than the mRNA data , we only reported events with at least two EST's supporting each isoform . This step was divided into two parts: the analysis of the core probesets , the roughly 240 , 000 probesets that target well supported exons , and the analysis of all probesets , which includes many more potential genes and gene regions with little evidence of ever being expressed . In both cases the algorithm was the same but we modified the thresholds in order to take into account the greater amount of noise from the non-core probesets . The normalization , summarization and regression steps were described in an earlier publication [6] . Briefly , PLIER was used to normalize the data and summarize the probe scores into probeset scores and linear regression was used to correlate the expression level of each probeset with the genotypes of every known SNP within 50kB of the gene . Subsequently , for each gene ( meta-probeset ) , we identified the most significant probeset-SNP association and used this SNP to report a regression p-value and fold-change between homozygous genotypes for every probeset in the gene . For SNPs with no homozygous minor allele individuals , the expression fold-change between the two genotypes was doubled in order to estimate the predicted fold-change between homozygotes . To choose probesets as candidates for AS , we applied a threshold p-value for the linear regression and an absolute value threshold for a score U , as described in Equation 1 , which represents the unexpectedness of the fold-change in the context of the other probeset expression changes in the gene . ( 1 ) , where , in log2 values , FCps is the probeset fold-change , FCgene is the average fold-change of all other probesets in the gene , σ is the standard deviation of the other fold-changes and K is a small constant which we arbitrarily set at 0 . 2 , added to control for genes with too little variation , or in other words , stabilize the variance . For probesets that passed both thresholds and had prior AS evidence , we plotted the log10 of the p-values and the log2 of the fold-changes of all probesets in the gene onto the UCSC Genome Browser as custom tracks and selected candidates for PCR validation via visual inspection . Links to all selected candidate tracks are available in Table S1 . The “AS-marker” track indicates the position of the affected probeset . We visually inspected 100 candidates from the analysis of core probesets and 160 candidates from the analysis of all probesets , in order to select only the ones where the microarray based and EST-based predictions where concordant in the context of the entire transcript . For the visually verified candidate probesets , we re-applied the linear regression using the probe-level scores and retained only the candidates for which a majority of individual probes agreed with the probeset-level regression ( which is based on PLIER's probeset summary scores ) . The unexpectedness ( U ) measure described in Equation 1 is better suited at finding short AS events , such as single cassette exons or alternative splice site usage rather than large splicing changes , such as the one observed in the ELAVL1 gene in the previous paper [43] , where half of the probesets can be included or excluded . This later case would cause the standard deviation of fold-changes to be very large , thus bringing down the score . We were specifically looking for short AS events since they are more readily amplifiable via PCR and hence easy to validate . The presence of the standard deviation in the denominator also helps to avoid noisy genes due to low expression or inconsistent splicing patterns , such as immunoglobulin genes , and false positives caused by the erroneous inclusion of multiple genes in the same meta-probeset . We selected 72 potential AS events for PCR validation . Although we had all 57 lymphoblast samples in our lab's possession , given limited resources , we selected a strategic sample of 10 individuals which together harbored variation for all 72 associated SNPs and forwarded the samples for validation of the alternative splicing events using semi-quantitative PCR . Primers were successfully designed for 62 events , but 4 of the reactions did not yield any product ( primer failure ) . The sizes and relative quantity of amplicons were determined using a Caliper LC90 automated chip-based capillary electrophoresis instrument and then sizes were matched-up with expectations based on AceView gene annotations , as detailed in the paper describing this automated AS validation procedure [44] . For the 58 reactions which produced viable results , we performed linear regression on the estimated ratios of each amplicon with the genotypes of the associated SNP . When the direction of the slopes and the lengths of isoforms agreed with the micro-array regression and the smallest regression p-value was below 0 . 05 , we considered these cases fully validated as genetically cis-regulated isoform level variations . If the lowest p-value was above 0 . 05 but there was a clear visible trend in the right direction , we classified them in the second category: not statistically significant , but two isoforms were present with a trend in the right direction . We also designated lower levels of validation confidence for cases that showed isoform level variations but where the association with the SNP could not be verified , and finally for cases that showed AS without detectable inter-individual variation . We performed sequencing of the genomic DNA of the regions around the 72 validation candidate AS events as well as 10 candidate events from previous analyses . We selected 4 individuals to sequence for every AS candidate exon . We chose 2 individuals to represent each distinct homozygous genotype and made sure they also represented the difference in probeset expression associated with the genotype . Heterozygotes were used in cases where the rare minor genotype was not present in our sample . We designed primers with Primer3 [45] to amplify a region of 600–800 bases centered on the probeset and sequenced these regions in the 4 individuals . In total , we retrieved useful sequence information for 60 of the regions and we were able to detect 86 SNPs . We used Phred [46] , Polyphred [47] and Consed [48] to analyze the chromatograms and we aligned the 4 output sequences to a reference sequence using ClustalW [49] to map the detected SNPs to the genome . To view the SNPs in the context of the gene structure and identify potentially important polymorphisms , we displayed the SNP positions with their associated micro-array expression change on the UCSC Genome Browser as custom tracks . These tracks are available in the Table S2 . Whenever a SNP was close enough to a splice-site , we used the MaxEntScan [15] online tool to score the theoretical strength of the different versions of the splice-site sequence . We found 13 SNPs that fell inside MaxEntScan's 9 base window around the 5′ splice-site and 3 SNPs that fell inside the 23 base window around the 3′ splice-site ( see Table 2 ) . In every case , we could show a qualitative agreement between the expected change in splice-site strength calculated by MaxEntScan and the micro-array expression change . For every SNP that was sequenced in an exon , we used the online tool ESE Finder 3 . 0 to predict potential exonic splicing enhancers ( ESEs ) that would be affected by the SNP ( see Table 3 ) . We tabulated the sequences and matrix-based scores of all ESE predictions containing the polymorphic base and surpassing ESE finder's established thresholds , except for the one ESE detected in the WARS gene , which was only 0 . 04 below the default threshold . We made use of in vivo minigene assays to verify the causal link between 6 candidate SNPs and detected AS events . The minigene assays were based on an article by Singh and Cooper [18] . We performed the experiment on 2 genes from this analysis , MMAB , and AMACR , and 4 top-scoring hits from the previous study , PARP2 , ERAP2 , ATP5SL and CAST . Primers were designed to amplify the alternatively spliced exon and 200bps from flanking introns . Genomic DNA was amplified for four individuals in each case , two with each homozygous genotype for the SNP of interest . Sequences were digested with Xba1 and Sal1 , as was the RHCGlo plasmid vector . The plasmid and amplicons were then ligated . Clones were purified , PCR-amplified and sequenced , to verify that the final plasmid sequences assayed differed only at the expected SNP . Then 2ug of the new plasmids were transfected into wells containing ∼3×1015 HeLa cells . 24–48h after transfection , RNA was extracted using Trizol , following the manufacturer's instructions . Alternative splicing was assessed via RT-PCR using the plasmid-specific primers RSV5U and TNIE4 [18] . Isoform presence was assessed by gel electrophoresis ( see Figure 3 ) and , in 3 cases ( AMACR , MMAB , ATP5SL ) quantified by capillary electrophoresis using the Agilent DNA 1000 Chip Kit , according to the manufacturer's protocol . We used the publically available program fdist2 [19] to calculate the Fst of all 41 HapMap SNP we sequenced based on the genotype frequencies in 4 populations , CEPH , YRI , CHB and JPT , merging the Asian population ( CHB and JPT ) frequencies together , as recommended by the fdist2 instructions . In order to estimate the percentile ranking of our SNP Fsts , we used fdist2 to create a simulation dataset of 20 , 000 data points based on 3 demes , 3 sample populations and an expected Fst of 0 . 187 , which was the average Fst estimated from 800 random HapMap SNPs . For all 41 re-sequenced HapMap SNP , based on the phased haplotypes of the CEU HapMap population ( Build 36 ) and a core haplotype of a single SNP , we measured the relative extended haplotype homozygosity , as defined by Sabeti et al . [49] , at every size below 500 kb , each step extending the haplotype by one SNP on each side . We reported the highest measured REHH for any block size , which was generally situated between 100 and 400 Kb . The minor core haplotype frequency had to be above 0 . 2 because lower frequencies can cause artifactually high REHH values [50] . We also insisted that REHH measurements be supported by at least 10 identical phased haplotypes to avoid performing the measurements on overly fragmented haplotypes . We performed the same analysis on 300 randomly chosen HapMap SNPs that fell within a RefSeq gene and used this distribution to estimate the relative percentile of the maximum REHHs of our sequenced SNPs .
|
Alternative splicing ( AS ) , through the alternative use of exons , can produce many different mRNA transcripts from the same genomic locus , thus possibly resulting in the production of many different proteins . We know that splicing differences between individuals exist and that these changes are often associated with genetic variants . Thus far , very few of these associations have led to the precise localization of the causative polymorphisms . In this work , using in-depth analysis of previously published splicing sensitive micro-array data from human cell lines , we identified and validated a large number of splicing changes which are highly correlated with nearby genetic variations . We then sequenced the genomic DNA around candidate exons and used in silico modeling tools to identify causative SNPs for most of our candidates . Using a plasmid reporter construct , we further demonstrated that five selected SNPs reproduce the expected effect in vivo . Our results indicate that genetically controlled splicing differences between individuals may be more common than previously suggested and can be very subtle; and most are caused by SNPs affecting either the splice-site region or exonic splicing enhancers ( ESEs ) sequences .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"molecular",
"biology/rna-protein",
"interactions",
"computational",
"biology/molecular",
"genetics",
"computational",
"biology/alternative",
"splicing",
"molecular",
"biology/rna",
"splicing",
"molecular",
"biology/bioinformatics",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"computational",
"biology/genomics"
] |
2009
|
Fine-Scale Variation and Genetic Determinants of Alternative Splicing across Individuals
|
Infections with Taenia solium are the most common cause of adult acquired seizures worldwide , and are the leading cause of epilepsy in developing countries . A better understanding of the genetic diversity of T . solium will improve parasite diagnostics and transmission pathways in endemic areas thereby facilitating the design of future control measures and interventions . Microsatellite markers are useful genome features , which enable strain typing and identification in complex pathogen genomes . Here we describe microsatellite identification and characterization in T . solium , providing information that will assist in global efforts to control this important pathogen . For genome sequencing , T . solium cysts and proglottids were collected from Huancayo and Puno in Peru , respectively . Using next generation sequencing ( NGS ) and de novo assembly , we assembled two draft genomes and one hybrid genome . Microsatellite sequences were identified and 36 of them were selected for further analysis . Twenty T . solium isolates were collected from Tumbes in the northern region , and twenty from Puno in the southern region of Peru . The size-polymorphism of the selected microsatellites was determined with multi-capillary electrophoresis . We analyzed the association between microsatellite polymorphism and the geographic origin of the samples . The predicted size of the hybrid ( proglottid genome combined with cyst genome ) T . solium genome was 111 MB with a GC content of 42 . 54% . A total of 7 , 979 contigs ( >1 , 000 nt ) were obtained . We identified 9 , 129 microsatellites in the Puno-proglottid genome and 9 , 936 in the Huancayo-cyst genome , with 5 or more repeats , ranging from mono- to hexa-nucleotide . Seven microsatellites were polymorphic and 29 were monomorphic within the analyzed isolates . T . solium tapeworms were classified into two genetic groups that correlated with the North/South geographic origin of the parasites . The availability of draft genomes for T . solium represents a significant step towards the understanding the biology of the parasite . We report here a set of T . solium polymorphic microsatellite markers that appear promising for genetic epidemiology studies .
Cysticercosis is an infection caused by the larval stage of the cestode Taenia solium . When the larval stages infect the central nervous system , the infection is known as neurocysticercosis ( NCC ) and is the most common cause of adult-onset seizures in endemic regions worldwide . Crude estimates of the burden of infection and disease suggest that greater than ten million people have NCC and as many as 2 . 7–5 . 6 million suffer from epilepsy [1] . A recent analysis concluded that in Latin America , vast parts of Asia , the Indian subcontinent and Southern China , Sub-Saharan Africa , and Oceania , 29% of all cases of epilepsy are attributable to NCC [2] . Humans are the only known definitive host , harboring the adult tapeworm and releasing infectious eggs to the environment [3] . In pigs that ingest infectious ova or proglottids , the released oncospheres cross the intestinal wall into the circulatory system where they become trapped in the microcapilaries , often in the brain , muscles and subcutaneous tissues . The oncospheres develop into cysticerci ( cysts ) and if present in the parenchyma of the brain , seizures and epilepsy may occur as a result of host inflammation against the cysts . Understanding the genetic variation of T . solium has the potential to improve our knowledge of the biology , epidemiology , infectivity , and pathogenicity of this parasite in endemic regions [4–7] . Moreover , analysis of the genetic variation within and between different geographical populations can provide information on evolution [8] , genetic differentiation and speciation of parasites [9] , as well as provide tools for understanding transmission dynamics , which may contribute to public health efforts to control this parasitic infection . The first attempts at genotyping Taenia parasites were directed towards the differentiation of Taenia species based on the sequence polymorphism of mitochondrial NADH dehydrogenase 1 and cytochrome c oxidase subunit I ( COI ) genes using single-strand conformation polymorphism ( SSCP ) [10] . Restriction fragment length polymorphism ( RFLP ) also was used to discriminate Taenia species by analyzing the ribosomal 5 . 8S gene sequence as well as the internal transcribed spacer ( ITS ) [11] . In 2001 , Hancock et al showed diversity among T . solium cysts from different countries using COI , a portion of the ITS1 encoded gene and the diagnostic antigen Ts14 . Little genetic diversity within T . solium samples collected from South America and Asia was observed . In addition , 15 isolates from Peru had similar COI sequences showing no genetic variability between them [12] . Later , two different worldwide genotypes were reported , with Asian parasites grouped into one cluster , and parasites from Latin America and Africa grouped into another cluster [4] . It has been suggested that the low variation found in T . solium isolates may be associated with the limited resolution of the experimental techniques used at that time [8] . More recently , with the development of new DNA analysis tools such as Random Amplification of Polymorphic DNA ( RAPD ) , greater genetic variation has been reported in parasites from communities in Mexico , Honduras and Madagascar [5 , 13–15] . This data suggests that T . solium has local lineages with different genetic characteristics . However , some disadvantages have been reported with RAPD such as low reproducibility , inability to test heterozygosity and subjective interpretation of the data [16 , 17] . Therefore , a more robust tool with higher resolution is needed to obtain more precise genotyping of T . solium isolates . Microsatellites , or Simple Sequence Repeats ( SSR ) , are repetitive DNA sequences consisting of blocks of 1 to 6 nucleotides repeated up to 60 times 8 . They are highly polymorphic in the number of repeated units . The variation in size of repeat domains is mainly generated by slippage of DNA polymerase during DNA replication , resulting in the insertion or deletion of one or more repeated units [18 , 19] . Microsatellites have the advantage of being multi-loci and principally neutral markers [19] , meaning that unlike protein-encoding genes , they are less likely to be subject to selective pressure . Microsatellites are highly reproducible and specific , and are easily identified from genome sequences by bioinformatics data mining [20–22] . Microsatellite polymorphisms can be detected by polymerase chain reaction ( PCR ) amplification followed by DNA electrophoresis [8 , 23] . This technique has been used to analyze genetic variation of other parasites such as Leishmania spp . [24] , Schistosoma japonicum [6] , Trypanosoma cruzi [25] , and Plasmodium falciparum [26] . Microsatellite markers also have the advantage of being able to detect greater genetic variation than other genetic markers , as has been demonstrated in Echinococcus multilocularis [27] . This work suggests that microsatellite polymorphism analysis is an appropriate tool to differentiate T . solium isolates . With the availability of draft genome sequences , the identification of microsatellites is more efficient [20] . Recently , a draft genome of a T . solium isolate recovered from Mexico has been published [28] . It is however necessary to have more genomic information available , in order to identify genotyping markers . In this study we present two draft genome sequences corresponding to T . solium specimens from Huancayo ( cysts ) and Puno ( proglottid ) from which we identified and characterized microsatellite markers . We explore microsatellite length variability to differentiate T . solium isolates from two regions of Peru . To analyze the microsatellites length we used a multi-capillary electrophoresis QIAxcel system that has the advantage of automatized size determination [29] . Although its advantages , this technique is limited by 3–5bp resolution that will not let us differentiate length polymorphism lower than 3–5 bp . The proposed microsatellites will lead to a more comprehensive understanding of the epidemiology of this important human pathogen .
Individual cysticerci ( cysts ) were recovered from a single , naturally infected pig from Huancayo , a city in the central Andean region . One proglottid from Puno ( Fig 1 ) was excised from segments of an adult tapeworm recovered from a single fecal specimen and used for extraction of DNA . To minimize contamination with exogenous materials , the proglottid was washed thoroughly 10 times with phosphate buffer solution ( PBS ) , transported in a mixture of PBS and antibiotics penicillin/streptomycin/ amphotericin B ) , and stored at -70°C until the DNA was extracted . For the assembly , both the Illumina mate-pair and paired-end sequencing results were combined , providing a total of 878 , 340 , 445 usable reads at 100 bp average length . In order to provide the best assembly at an average of 157X coverage and for an expected genome size of 115MB , a subset of 175 , 931 , 369 reads were selected for de novo assembly process using Velvet v1 . 1 . 05 [30] , with the following parameters: coverage cutoff = 10 , expected coverage = 26 , paired end insert length = 350 , and mate paired insert length = 3 , 500 . For the Puno proglottid genome assembly , both the Illumina mate-pair and the 454 GS FLX raw signals data were processed with the software GS Run Processor to obtain the reads , which were assembled using GS de novo Assembler 2 . 6 with the following parameters: minimum read length = 20 nucleotides ( nt ) , overlap seed step = 12 nt , overlap seed length = 16 nt , overlap minimum match length = 40 nt , overlap minimum match identity = 90 nt , overlap match identity score = 2 nt , overlap match difference score = -3 nt , all contig threshold = 100 nt , large contig threshold = 500 nt; with an expected depth of 25X . Duplicated reads were used and the assembly was performed 30 times , taking the iteration with the mean number of contigs . To generate a more complete T . solium genome sequence , we produced a hybrid assembly consisting of the Puno-proglottid fragment 454 reads combined with the Huancayo-cyst Illumina paired-end reads as follows . The hybrid genome was generated using Velvet v1 . 1 . 05 de novo assembler using the Puno-proglottid 454 fragment reads , Huancayo-cyst Illumina paired-end reads , and Huancayo-cyst Illumina mate-pair reads . Velvet parameters used are as follows: kmer value = 57 , coverage cut-off = 9 , expected coverage = 38 , paired insert length = 350 , and mate pair insert length = 3500 . The set of contigs corresponding to the genome of the Huancayo-cysts , the set of contigs corresponding to the assembled genome of the Puno-proglottid and the set of contigs corresponding to the hybrid assembly created from both the Huancayo and Puno tissue genome assemblies is available together at Genbank under the project ID PRJNA183343 . Microsatellites were identified using the script developed by Gur-Arie et al [31] . We searched for repetitive motifs of 1–6 bp in our two assembled T . solium genomes . A total of 36 distinct microsatellites with polymorphic attributes were selected according to the following criteria: For the Puno-proglottid genome , a first group of microsatellites with a minimum of five motif repetitions were selected . In order to determine if the microsatellites are transcribed , we verified if they were present in the T . solium ESTs sequences database available ( http://www . ncbi . nlm . nih . gov/nucest/ ? term=%22Taenia+solium%22%5Bporgn%3A__txid6204%5D ) , which later was included in the published T . solium genome sequence [28] . In order to determine if the microsatellites are comprised within a coding region , we verified the presence of an ORF using the algorithm ORF-Finder ( http://www . ncbi . nlm . nih . gov/gorf/gorf . html ) . Microsatellites that showed differences in the number of repeats between the Puno-proglottid genome and the ESTs database were selected for further analysis . We verified that conserved flanking regions were present in both the Puno-proglottid genome and the ESTs database . Five microsatellites ( TS_SSR01 to T S_SSR05 ) were selected . In addition , 8 of the longest microsatellite sequences from the Puno-proglottid genome not present in the ESTs database also were selected ( TS_SSR06 to TS_SSR13 ) . Because this first group of microsatellites was biased based on their availability in ESTs , it is less likely that they are neutral . A second group of microsatellites was identified in the Puno-proglottid genome and mapped into the corresponding contigs of the Huancayo cyst genome . 200 microsatellites ( 100 di-nucleotides and 100 tri-nucleotides ) with the largest repetitive motif in both genomes were evaluated . After aligning the sequences we selected the largest sequences that showed polymorphisms in the repetitive sequences between both genomes ( 13–18 repeats ) . This resulted in additional 23 microsatellite loci for testing . Taenia solium cysts were excided in a previous study from a naturally infected pig in a Huancayo local abattoir . The pig was bought by the study team at market price so the study team owned the animal . Procedures were approved by Universidad Peruana Cayetano Heredia ( UPCH ) ethics committee for animal use T . solium proglottid specimens were collected in previous studies by the Cysticercosis Working Group in Peru , with approval of the UPCH IRB ( IRB00001014 ) ; they were used as residual diagnostic samples .
Sequencing of the Huancayo-cyst genome produced a total of 175 , 931 , 369 reads that were assembled into 18 , 361 contigs ( the largest contig was 307 , 365 bp ) . The estimated genome size was 114 , 605 , 177 nt . The sequencing of the Puno-proglottid genome produced a total of 76 , 625 , 473 reads that were assembled in 47 , 475 contigs ( the largest contig was 79 , 438 nt ) . The lack of large contigs in this assembly is due to the lack of paired-end sequencing data . The estimated size of the Puno-proglottid genome was 109 , 898 , 809 nt . For the hybrid cyst/proglottid assembly , 7 , 979 contigs were obtained , and the largest contig had 395 , 362 nt . The estimated size of the hybrid genome was 111 , 029 , 218 nt . These and other statistics were calculated with the program multifastats , py v1 . 4 ( https://github . com/lbbm-upch/multifastats_v1 . 4 ) , and are summarized in Table 2 . As expected due to their closeness , the Peruvian T . solium genomes showed a similar size as the recently published Mexican T . solium genome ( 122 . 3 Mb ) , as well as the genomes of E . granulosus ( 114 . 9 Mb ) , and E . multilocularis ( 115 Mb ) [28] . We identified 9 , 129 microsatellite sequences distributed in the T . solium Puno-proglottid genome and 9 , 936 in the Huancayo-cyst Genome ( Table 3 ) . In both the Puno and Huancayo genomes , the greatest number of microsatellite loci found contained di-nucleotides repeats . Most of these microsatellites were over-represented in the forms of AC/GT and AG/CT , while the forms AT/AT and CG/CG showed a lower frequency of occurrence ( S1 Table ) . Thirty-six microsatellites markers were identified as potentially polymorphic . We successfully amplified 34 microsatellite markers in 40 T . solium tapeworm specimens . The estimated size of the PCR products of the microsatellites in the tapeworm isolates was similar to the expected theoretical sizes . Within the tested sample , twenty-seven microsatellite markers were monomorphic , of which 26 were homozygous ( only one band was observed in the electrophoretic pattern ) and 1 was heterozygous ( TS_SSR31 , in which two bands were observed in all samples ) . Seven microsatellite markers were polymorphic containing a total of 44 alleles within the polymorphic loci ( Table 4 ) . All 40 tapeworms were homozygous for five markers ( TS_SSR09 , TS_SSR16 , TS_SSR18 , TS_SSR27 , and TS_SSR28 ) , while some were heterozygous for TS_SSR01 and TS_SSR32 ( S2 Table ) . The number of alleles varied from 4 for locus TS_SSR16 to 10 for locus TS_SSR32 with an average of 6 alleles per locus . The polymorphic information content ( PIC ) varied from 0 . 472 for locus TS_SSR16 to 0 . 843 for locus TS_SSR28 with an average of 0 . 604 per locus ( Table 4 ) . The reproducibility analysis of TS_SSR01 amplification showed a variability of 1–2 bp between the four replicas in the nine isolates tested ( S3 Table ) , which is lower than the range of resolution reported by the manufacturer ( 3–5 bp ) . The variability of TS_SSR01 size between the different T . solium isolates appeared as 2–15 bp , which is three fold higher than the experimental error reported by the manufacturer ( S3 Table ) . Only TS_SSR01 was found to be present in the EST database . However the comprising region did not show evidence of any ORF . Therefore TS_SSR01 is part of a transcribed but not-translated sequence . When we compared TS_SSR01 against the complete no-redundant nucleotide collection of Genbank using blastn , only one sequence from the close organism T . asiatica appeared similar . The genetic diversity observed in isolates from the southern city of Puno ( 20 different genotypes ) was slightly higher than that the genetic diversity observed in the isolates from the northern city Tumbes ( 16 different genotypes ) . Also the median number of alleles was slightly higher ( Table 5 ) . The single microsatellite that best differentiated tapeworms from Tumbes and Puno isolates was TS_SSR01 . Most of the isolates from Tumbes ( 17/20 ) were associated to genotype A ( 206/206 bp ) and 3/20 of isolates were associated to genotype B ( 211/211 bp ) , all of them homozygous . A lower prevalence of genotype A was observed in Puno ( 6/20 ) compared to Tumbes ( P = 0 . 001 , Fisher’s exact test ) . A similar prevalence of genotype B ( 2/20 ) was observed in Puno . Five other genotypes were found only in Puno ( S2 Table ) .
The present study describes the draft genome sequences of two T . solium isolates and the identification and characterization of DNA microsatellites . The T . solium microsatellites reported here were found to be distributed along the entire genome . The length polymorphism of microsatellites was analyzed for its association with the geographic origin of tapeworm isolates . We found novel microsatellites that were able to differentiate tapeworms between the northern and southern regions of Peru . Microsatellites have proven to be highly informative in population genetic studies in several parasites [6 , 33] . In the particular case of T . solium , the use of microsatellite markers allows a way to define the genetic structure of populations and to conduct genetic epidemiology studies . Although previous studies have shown a moderate genetic diversity of T . solium [4 , 9 , 10 , 34] and particularly in Peru [12] , the novel microsatellites we identified here have demonstrated the capacity to differentiate tapeworms from Tumbes in the north and Puno in the south of Peru . Although the frequency of microsatellites and their coverage in the genome varies considerably between organisms , the number of microsatellites found in T . solium ( between 9 , 000–10 , 000 ) is similar to the number of microsatellites identified in other parasites [21 , 35] . Although most of the microsatellite sequences were found in non-transcribed regions , we found that T . solium microsatellites could also be present in transcribed/non-translated regions , being the abundance in non-transcribing regions higher than in transcribed/non-translated regions . This result is consistent with previous studies that reported that microsatellites are more abundant in non-coding regions of eukaryotic organisms [36] . The relatively low abundance of microsatellites in transcribed/non-translated as well as in coding regions could be explained by a negative selection against mutations that change the function by altering the secondary structure of the transcribed sequence or by altering the reading frame of the genes [36 , 37] . Eukaryote microsatellite loci typically contain between 5 and 40 repeats , similar to what we found in T . solium . As in other organisms , the number of microsatellites in T . solium decreases as the size of the repeat unit increases [38 , 39] . It is important to highlight that the distribution of the repeat types ( mono- to hexa-nucleotide ) varies across different taxa , and it has been suggested that this variation is associated with to the interaction of the mutation and the differential selection pressure [37] . As previously reported in other species , we found that dinucleotide repeats motifs were the most abundant type in T . solium , which tend to be longer in non-coding regions . This seems to be explained by the negative selection pressure of polymerase slippage during replication of coding DNA [40] . Castagnone—Sereno et al . reported that in nematodes , ( AT ) n was the most common microsatellite motif [35] . We found the AC / GT dinucleotide motif to be the most abundant in T . solium , which concordantly has also been found to be common in most vertebrates and arthropods [41] . The genetic variability observed in this study may be explained by several factors , including migration of humans and pigs , mutations in the tapeworm genome , cross-fertilization of tapeworms in the intestine in cases where multiple tapeworm infections occur [42 , 43] , among others . It is important to note that although the low resolution of QIAxcel system reported by the company ( 3–5bp ) , the range of difference in the size of TS SSR01 between the north and south region of Peru , is 2–3 fold higher than the expected error ( 5–15 bp ) and the results of the repeatability assay showed lower variability ( 1–2 bp ) . This evidence supports the main finding of having TS SSR01 as a polymorphic marker able to differentiate tapeworms from the north and south region of Peru . Transmission dynamics are not fully understood , although genetic characterization by means of microsatellite genotyping may unveil details of the ecology of T . solium . The use of molecular characterization by means of microsatellites will potentially allow identification of genetic links between tapeworms , larval cysts found in infected pigs and eggs in soil or fomites . Furthermore , microsatellites would help disentangle the genetic complexity of a population due to the introduction of external tapeworms from immigrant tapeworm-carriers . This method of genotyping also has implications in the evaluation of parasite control by identifying the source of infection , and the re-introduction routes of the parasite into a specific region . In conclusion , this study describes the identification and application of microsatellite markers in T . solium genotyping . The novel microsatellites reported here would be an important tool for future studies of the genetic variability of T . solium , including population genetics , basic epidemiology , super infections with more than one strain , and tracking the transmission of cysticercosis .
|
Taenia solium , the pork tapeworm , is an important pathogen as it is a major cause of acquired epilepsy in developing countries . The parasite was eliminated from most developed countries decades ago due to improvement in sanitary conditions but it remains a common infection across Asia , Africa and Latin America . Identification of genetic variants within T . solium will enable to study the genetic epidemiology , distribution and movement of this parasite within endemic communities , which will ultimately facilitate the design of control strategies to reduce the health and economic burden of disease . Microsatellites have been used in other parasites to identify genetic variants . In this study , we partially sequenced the genome of T . solium and identified microsatellites widely distributed in the genome using bioinformatics tools . We evaluated the distribution of these microsatellites collected from 20 tapeworms from the north and 20 tapeworms from the south of Peru . We identified seven polymorphic microsatellites , and evaluated their capacity to differentiate genetic variants of T . solium . Interestingly , tapeworms from the North and South of Peru showed different genotypes , suggesting its use as a potential marker to differentiate geographic origin .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Identification and Characterization of Microsatellite Markers Derived from the Whole Genome Analysis of Taenia solium
|
The rearrangement of protein domains is known to have key roles in the evolution of signaling networks and , consequently , is a major tool used to synthetically rewire networks . However , natural mutational events leading to the creation of proteins with novel domain combinations , such as in frame fusions followed by domain loss , retrotranspositions , or translocations , to name a few , often simultaneously replace pre-existing genes . Thus , while proteins with new domain combinations may establish novel network connections , it is not clear how the concomitant deletions are tolerated . We investigated the mechanisms that enable signaling networks to tolerate domain rearrangement-mediated gene replacements . Using as a model system the yeast mitogen activated protein kinase ( MAPK ) -mediated mating pathway , we analyzed 92 domain-rearrangement events affecting 11 genes . Our results indicate that , while domain rearrangement events that result in the loss of catalytic activities within the signaling complex are not tolerated , domain rearrangements can drastically alter protein interactions without impairing function . This suggests that signaling complexes can maintain function even when some components are recruited to alternative sites within the complex . Furthermore , we also found that the ability of the complex to tolerate changes in interaction partners does not depend on long disordered linkers that often connect domains . Taken together , our results suggest that some signaling complexes are dynamic ensembles with loose spatial constraints that could be easily re-shaped by evolution and , therefore , are ideal targets for cellular engineering .
Cell signaling networks possess a remarkably modular organization . This modularity has attracted the attention of synthetic biologists , for it offers a plausible approach to engineer novel and useful cellular behaviors . At the center of this modular organization are protein domains , which are recurring structural units that often perform modular and thus portable functions [1] , [2] . In most signaling proteins , multiple domains are connected by flexible linkers [3] . Diverse genetic mechanisms can rearrange domains , leading to the creation of proteins with novel domain combinations [4]–[8] , and thus altered functions [9]–[11] . While , the most prevalent mechanism is gene duplication and in frame fusion , followed by the loss of terminal domains [12] , other mechanisms such as transpositions , translocations , inversions , or recombinations , though less prevalent , can also lead to the same functional outcome [4] . Experimental [9] , [11] and computational [5] , [10] , [13]–[15] efforts have revealed that domain rearrangements play a prominent role in the mutational re-wiring of signaling networks , with clear consequences for evolution [4] , [12] and disease [16] . Furthermore , the versatility conferred by the functional modularity of protein domains , has begun to be harnessed by protein engineers and synthetic biologists [17]–[19] , and promises to open new avenues for cellular engineering [20] . While in nature domain rearrangements can occur by a variety of mechanisms [4] , [5] , [7] , [8] , [12] , in principle two major outcomes are possible: a protein with a new domain combination is created , without altering pre-existing genes ( Figure 1A ) , or the creation of a protein with a new domain combination concomitantly replaces a protein with a pre-existing domain combination ( Figure 1B ) . Recently , it has been shown that signaling network function can be altered , when domain-rearrangement events create proteins with new domain combinations without replacing pre-existing genes [11] . While this work demonstrated that domain rearrangements could be a major force in the evolutionary diversification of signaling pathways , a far more challenging question still needs to be addressed: How are domain rearrangements tolerated when the genetic mechanisms involved result in the simultaneous replacement of a pre-existing gene ? Understanding how these replacements are tolerated is difficult if one considers that not all of the pre-existing functions are preserved , and that the proteins involved are often part of large multi-protein complexes believed to have defined 3D structures , and thus likely to impose spatial constraints . Furthermore , it has been postulated recently that , because of the possible combinatorial complexity involved in the assembly of multi-protein complexes , signaling complexes within a cell might have compositional heterogeneity [21]–[23] . Thus , to fully comprehend how domain-rearrangements may affect signaling networks function , it is also necessary to understand the mechanisms by which the concomitant replacements are tolerated . This knowledge would advance our understanding of fundamental aspects of network evolution and , as importantly , could enable the development of efficient tools for signaling engineering . To answer this question , we used a synthetic biology approach to systematically determine the robustness of the mitogen activated protein kinase ( MAPK ) -mediated yeast mating pathway ( described in Figures 1B and S1 ) [24] to 92 domain-rearrangement events that replace 11 pre-existing genes . Specifically , we created a library in which 22 domains from 11 mating pathway proteins were shuffled ( as described in [11] and shown in Figure 1C ) . Library variants in which domains from a given protein ( e . g . , the MAP3K Ste11 in the example in Figure 1D ) were shuffled with domains from all other proteins , were transformed into a yeast strain in which the corresponding gene ( Ste11 in the example ) had been previously deleted . In this manner , we effectively replaced a wild type ( WT ) gene with all library constructs that include at least one domain from the replaced gene . By repeating this procedure for individual deletion strains in which either the G protein β subunit Ste4 , the G protein γ subunit Ste18 , the scaffold Ste5 , the adaptor Ste50 , the PAK kinase Ste20 , the MAP3K Ste11 , or the MAP2K Ste7 had been deleted , we generated a library of strains in which a WT gene has been replaced by all domain-rearrangement mutants involving domains from that gene ( Figure S2 ) . The mutational mechanisms that rearrange domains in natural proteins are obviously different from the two-part shuffling method used to construct our library . However , we are not interested here in investigating specific mechanisms leading to domain rearrangement , but rather the functional consequences that these rearrangements have at the protein and network level . Furthermore , while evolution could rearrange domains from any pair of proteins in the genome ( though recent evidence suggests that rearrangements can preferentially occur among functionally related genes [25] ) , for simplicity we limited our analysis to rearrangement events between proteins belonging to the mating pathway . Assessing how general the results presented here are would require a genome-wide analysis that is beyond the scope of this work .
To determine how domain-rearrangement mutations that replace pre-existing genes affect network function , we measured by flow cytometry the fluorescence levels of a green fluorescent protein ( GFP ) reporter controlled by a mating-responsive pFUS1 promoter , before and 2 hours after stimulation with 1 µM mating pheromone ( Figure 2A ) . As a control , we first confirmed that deletion of each individual gene abolishes pathway activation ( with the exception of Ste50Δ that can still mediate very low , though statistically significant pathway activation , t-test , p≤0 . 013 ) ( Figure 2B ) . Remarkably , we observed that 34 out of the 92 tested domain-rearrangement variant strains rescued pathway activation in a pheromone-dependent manner , above the levels observed in the corresponding deletion strains ( Figures 2C and S3 , with statistical analyses shown in Figure S4 ) . In addition to changes in gene expression , mating pathway activation induces polarized growth that results in cell-cell fusion . To determine whether the tolerated rearrangement-derived replacements could also mediate polarized growth and cell-cell fusion , we determined the presence of pheromone-induced shmoos by microscopy and performed also quantitative mating assays . As shown in Figure S5 , strains expressing active variants are capable of polarized growth . Furthermore , as shown in Figure S6 , about 75% of the tested variants mate with at least 10% of the WT efficiency , while among those , ∼30% mate as efficiently as WT ( for most variants , GFP-expression levels and mating efficiency correlate ) ( Figure S7 ) . Taken together , our results demonstrate that domain rearrangement-mediated replacements can be tolerated , in some instances with pathway activation levels indistinguishable from WT . While the mating pathway seems capable of tolerating domain rearrangement-mediated replacements , it is possible that at least some of these replacements could have detrimental effects on other cellular processes , and thus their evolutionary relevance would be questionable . To investigate this possibility , we measured the growth rate ( as a proxy for fitness ) of the most active domain rearrangement variants ( nine variants in four different deletion strains ) . As shown in Figure S8A , the growth rate of the domain rearrangement variants is equal to , or even higher than , the growth rate of the WT strain , indicating that , at least under the laboratory conditions tested , domain rearrangement does not affect fitness negatively . Furthermore , for the domain rearrangement variants that functionally replace the MAP3K Ste11 in the Ste11Δ strain , we also measured growth rate under high osmolarity-induced stress , as in addition to mediating the mating response , Ste11 is also a MAP3K in the high osmolarity pathway [26] . As shown in Figure S8B , growth rates are not negatively affected by domain rearrangements involving Ste11 , again suggesting that they do not impair fitness under the tested laboratory conditions . While it is likely that some domain combinations would be unable to fold and/or function properly , there is no simple correlation between a domain rearrangement variant expression level and its ability to mediate mating pathway response ( Figure S13 ) . More likely , analysis of the data in Figure 2C reveals that while some components of the signaling network are essential , other are interchangeable . The integrity of the Ste5 scaffold seems critical for pathway function , as domain-rearrangement events involving Ste5 domains are never tolerated . Similarly , kinase domains cannot be replaced even by other kinase domains ( e . g . , replacement of the MAP2K Ste7 kinase domain by those of the PAK Ste20 , MAP3K Ste11 , or MAPK Fus3 , failed to rescue pathway activity in the Ste7Δ strain ) . These results suggest that kinase-substrate specificities are firmly defined . In contrast , pathway function can be preserved when the N-terminal ( N-t ) interaction domains of the kinases Ste20 , Ste11 , or Ste7 , responsible for localizing the kinase domains to the signaling complex , are replaced with alternative interaction domains . Thus , we hypothesized that the ability of the network to utilize alternative mechanisms of kinase recruitment to the signaling complex may contribute to network robustness against domain rearrangement-mediated replacements . To explore this hypothesis , we first compared mating pathway activation mediated by the domain-rearranged kinase variants with those of kinase variants lacking N-t localization domains . As shown in Figure 3A ( and further analyzed in Figures S9 and S10 ) , variants lacking N-t localization domains activate pathway response very poorly , as compared to kinase variants with rearranged N-t domains . Second , we introduced in the N-t localization domains mutations that had been shown to reduce binding affinity with their respective interaction partners . Specifically , we mutated the following residues ( Figure 3B ) : I90K in Ste50's N-t SAM domain , known to reduce binding to Ste11's N-t SAM domain [27]; C177A and C180A in Ste5's RING domain , known to reduce binding to the Gβ Ste4 [28]; and H345D H348D in Ste20's PBD domain , known to reduce binding to the small GTPase Cdc42 [29] . As shown in Figure 3C , in seven out of eight cases , point mutations reduced pathway activation between 40%–60% , suggesting that the alternative N-t localization domains are effectively recruiting the kinases to the signaling complex . Finally , we further confirmed this hypothesis by fluorescence microscopy , using GFP-tagged domain rearranged variants . As depicted in Figure 3D , kinases with rearranged N-t interaction domains are still recruited to the mating shmoo , suggesting that they localize to the signaling complex ( note that when GFP is expressed alone , it is uniformly distributed in the cytoplasm ) ( Figure S11 ) . Thus , we conclude that the ability of the signaling complex to accommodate alternative mechanisms of kinase recruitment ( Figure S12 ) contributes to the robustness of the network to domain rearrangement-mediated replacements . We then investigated the mechanisms that enable the signaling complex to tolerate changes in recruitment interactions . Domain-domain interactions depend on specific binding interfaces , thus they are unlikely to tolerate drastic changes in interaction partners . Thus , if alternative recruitment has to maintain specific domain-domain interactions , two hypotheses are possible ( Figure 4A ) : ( i ) rearrangements are tolerated because , even though signaling complexes possess precisely defined spatial constraints that result in fairly rigid 3D structures , domains are connected by long and flexible linkers ( e . g . , Ste5 , Ste7 , Ste11 , Ste20 , and Ste50 are predicted to contain intrinsically disordered regions [IDRs] [30] ranging from ∼74 to ∼207 amino acids long , within their inter-domain linkers , see Figure S14 ) ; or ( ii ) rearrangements are tolerated because signaling complexes do not possess rigid spatial constraints , but are rather diffuse ensembles of dynamically interacting proteins [31]–[33] . We reasoned that , if signaling complexes had defined spatial constraints and thus IDRs were required for networks to tolerate domain-rearrangement-mediated replacements , deletion of IDRs located within inter-domain linkers should substantially reduce pathway function . In contrast , if signaling complexes do not possess tight spatial constraints , but are rather loosely defined regions in which multiple weak interactions create higher local concentrations of signaling proteins , then deletion of inter-domain IDRs should not be detrimental to pathway function . To differentiate between these two hypotheses , we deleted segments of 171 amino acids from Ste20's IDR , 97 amino acids from Ste11's IDR , and 74 amino acids from Ste50's IDR and determined the ability of the shortened proteins to mediate pathway activation , as compared to their respective full-length variants . Note that we did not analyze IDRs present in Ste7's N-t and Ste5's N-t or C-terminus ( C-t ) because they do not connect pairs of domains , and thus are not likely to facilitate inter-domain flexibility . As shown in Figure 4C , all three shortened variants are still capable of mediating pathway activation ( the decrease observed with Ste20's short variant is expected , as Ste20's IDR contains a proline-rich motif needed for proper binding of Bem1 , a Cdc42 interaction partner , that when mutated has been shown to reduce pathway activation by ∼50% [34] ) . To further explore the role of IDRs in pathway function , we simultaneously replaced two WT proteins for their corresponding shortened variants . As shown in Figure 4D , co-expression of IDR-deleted Ste11 and IDR-deleted Ste20 variants effectively mediates pathway activation in the double deletion strain Ste20Δ Ste11Δ; similarly , co-expression of IDR-deleted Ste11 and IDR-deleted Ste50 variants effectively mediates pathway activation in the double deletion strain Ste50Δ Ste11Δ . These results indicate that the mating signaling complex can tolerate simultaneous deletions of IDRs in at least two proteins . Finally , we asked whether IDRs were necessary to tolerate domain rearrangement-mediated replacements , by measuring pathway activation for IDR-deleted domain-rearranged variants , as compared to their respective full-length variants . As shown in Figure 4E , deletion of the IDRs does not reduce pathway activity for most of the domain rearrangement variants tested , suggesting that IDRs are not needed to tolerate domain rearrangement-mediated replacements . Taken together , our results suggest that the yeast mating signaling complex does not possess a rigid , precisely defined spatial geometry , or that at least multiple alternative conformations are functional . Though not the focus of this study , we also noticed that , in some instances , deletion of the IDRs increased basal levels of pathway activation ( Figure S15 ) . This observation suggests that , while IDRs are not required for pathway function , they might have regulatory roles . We hypothesized that if signaling complexes possess loosely defined spatial constraints , the network should tolerate more complex domain rearrangement events , such as those in which domains from pairs of proteins are reciprocally rearranged ( Figure 5A ) . To test this hypothesis , we introduced the pairs of reciprocally rearranged variants Ste20[N]-Ste11[C]+Ste11[N]-Ste20[C] , Ste7[N]-Ste11[C]+Ste11[N]-Ste7[C] , and Ste50[N]-Ste11[C]+Ste11[N]-Ste50[C] , in the double deletion strains Ste20Δ Ste11Δ , Ste7Δ Ste11Δ , and Ste50Δ Ste11Δ , respectively . As shown in Figure 5B , while transformation with any of the single domain rearrangement variants did not rescue the double deletions , transformation with each pair of reciprocally rearranged variants rescued pathway activation , demonstrating that the mating signaling complex can accommodate changes in domain connectivity in two components simultaneously , supporting the hypothesis that the signaling complex does not possess a rigid , precisely defined geometry . Some of the changes in network topology resulting from domain rearrangement events in our experiments mimic changes in network topology that have occurred during evolution . For instance , in yeast , the adaptor Ste50 mediates the interaction between the MAP3K ( Ste11 ) and the small GTPase ( Cdc42 ) ( Figure 5C ) . In contrast , in humans the adaptor Ste50 has been lost and , instead , there is a direct interaction between the MAP3K Raf and the small GTPase Ras [35]–[38] . The domain rearrangement variant Ste20[N]-Ste11[C] topologically resembles human Raf , as the N-t Ste20 PBD domain interacts with Cdc42 directly ( in Raf this interaction is mediated by the RBD domain , but Ste11's RBD binds Ste5 ) . To test the hypothesis that Ste20[N]-Ste11[C] could functionally resemble Raf , we measured the ability of the domain rearrangement variant Ste20[N]-Ste11[C] to mediate pathway activation in a strain in which both Ste50 and Ste11 had been deleted . As predicted , expression of Ste20[N]-Ste11[C] complements the simultaneous loss of Ste50 and Ste11 ( Figure 5D ) , confirming that the network topology evolved in our experiment functions similarly to the network evolved in metazoans . Finally , among the seven kinase-containing domain combinations that in our experiments resulted in active pathways ( Figure S16A ) , three have not been previously found in yeast mating pathway proteins ( e . g . , domain combinations in which Cdc42's small GTPase domain , Ste5's RING domain , or Ste4's β-propeller domain , are connected to kinase domains ) . However , as these domain combinations lead to functional proteins in our model system , we hypothesized that proteins with similar domain combinations are likely to be found in natural genomes . To explore this hypothesis , we searched the Domain Club Database [10] to identify natural proteins with domain combinations resembling those found in our experiments . As shown in Figure S16B , we identified the human proteins: ( i ) LRRK1/2 , with a domain composition that includes both small GTPase and kinase domains; ( ii ) PIK3R4 ( a regulatory subunit of the PI3K complex ) with a domain composition that includes both β-propeller and kinase domains; and ( iii ) MAP3K1 , with a domain composition that includes RING and kinase domains . Thus , while the functions of these proteins need not be related to those in our experiments , these results indicate that the novel domain combinations that lead to active proteins in our screening have also evolved naturally .
Our results indicate that the yeast mating pathway is remarkably robust to domain rearrangement-mediated replacements , tolerating multiple changes in recruitment interactions . In particular , we observed that the N-t domains or motifs of the three multi-domain kinases in the mating pathway ( Ste20 , Ste11 , and Ste7 ) , which normally localize the respective kinase domains to the mating signaling complex , can be replaced by alternative interaction domains ( from other kinases , or from other pathway components ) . In contrast , kinase domains cannot be replaced , suggesting that the specificity of kinase-substrate interactions is key for proper signaling function . Thus , while the inter-molecular connectivity of the domains is important , the intra-molecular connectivity is not as important , suggesting that proper network function depends more on the formation of a signaling complex composed of key domains , rather than key proteins . Interestingly , even though intra-molecular interactions between different domains within a protein may play regulatory roles [17] , we observed that for most domain rearrangement variants , the basal levels of pathway activation are similar to , or only slightly higher than those of the WT pathway ( Figure S3 ) . This may simply reflect the fact that activation of Fus3 , the bottom kinase in the pathway , requires two concurrent stimuli: ( i ) phosphorylation-dependent activation of the MAP2K Ste7 , and ( ii ) pheromone-dependent activation of the mating scaffold Ste5 [39] . Thus , even if domain rearrangement altered intra-molecular regulation and therefore increased the activity of upstream mating kinases , signal propagation would still depend on phosphorylation-independent activation of Ste5 . Because domain-domain interactions are highly specific , proteins with rearranged domain compositions may have altered localization within the signaling complex . How can then domain rearrangements be tolerated ? Initially , we hypothesized that the presence of long , disordered inter-domain linkers may enable each domain within a rearranged protein to localize to the correct site within the complex . However , we found that the IDRs present within inter-domain linkers are dispensable for pathway function and , more importantly , for the robustness of the network to domain rearrangements . Taken together , these observations suggest that the function of the mating signaling complex is not constrained by a defined geometry . Thus , we propose that , rather than a precisely assembled multi-molecular machine , the yeast mating signaling complex is an ensemble of dynamically interacting molecules with loose spatial constraints [40] , [41] . A “tridimensional meshwork , ” in which individual components are only transiently bound by multiple , weak interactions , makes sense if one considers that mating signaling complexes should be able to rapidly re-orient to follow changes in the direction of the pheromone gradient , as well as to accompany the growth of mating projections . Similar matrix-like meshworks have been postulated to explain the dynamic nature of microtubule plus-end tracking proteins , which rapidly track microtubules movement [42] . Furthermore , recently Mayer , Deeds , and their co-workers have postulated that , because of the combinatorial complexity involved in the assembly of multi-protein complexes , rather than a single complex with a defined composition , it is more likely that multiple complexes with different compositions might exist simultaneously [21]–[23] . Remarkably , Suderman and coworkers [23] computationally modeled the yeast mating pathway and showed that the mating signal could still be propagated by compositionally heterogeneous populations of complexes . Our results suggest that signaling complexes are not only compositionally heterogeneous , but also structurally flexible . While in the short term mutational robustness buffers the impact that genotypic changes could have on phenotype , in the long term , mutational robustness may facilitate evolution [43] , [44] . In particular , and as best described by Gerhart and Kirschner in their theory of facilitated variation [45] , mutational robustness may enable the network to explore regions of genotypic space that , though presently neutral , could lead to adaptation in the event of future environmental or genetic changes [46] , [47] . The relaxed spatial constraints of the mating signaling complex may enable the network to tolerate changes in protein interactions that result from the mutational events that lead to domain rearrangements . While still hypothetical , one could imagine that proteins with altered domain compositions may eventually evolve novel functions [10] . As affordable genome and transcriptome sequencing are rapidly expanding the list of domain-rearrangement mutations involved in disease [48] , our work may help understand how disease-causing mutations affect the function of signaling complexes with components homologous to those analyzed in this work . Finally , the fundamental principles revealed here suggest that flexible multi-protein complexes could be ideal targets for cellular engineering [20] .
Deletion strains were derived from a W303 strain with the following genotype: MATa , bar1::NatR , far1Δ , mfa2::pFUS1-GFP , his3 , trp1 , leu2 , ura3 . Seven strains were created in which the following genes from the mating pathway were deleted individually: Ste4 , Ste5 , Ste7 , Ste11 , Ste18 , Ste20 , and Ste50 , in all cases using Trp as a selectable marker . Deletion strains were validated by genomic PCR and flow cytometry ( each individual deletion of a “Ste” gene impaired pathway-dependent GFP expression ) . Double deletion strains ( Ste20Δ Ste11Δ , Ste7Δ Ste11Δ , and Ste50Δ Ste11Δ ) were also made by homologous recombination , using Leu as the second selectable marker . The domain-rearrangement libraries were designed and constructed using a previously described combinatorial cloning strategy [11] . All variants were expressed from centromeric plasmids with Leu selection , under control of a constitutive low expression promoter consisting of a 250-bp fragment of the CycI promoter , and an AdhI transcription terminator . Each strain carrying an individual deletion ( or a double deletion , as in Figure 5 ) was transformed with a domain-rearrangement variant ( or a combination of two , as in Figure 5 ) that effectively replaced the deleted gene ( s ) . Samples were induced with 1 µM α-factor ( Zymo Research ) , while controls were left untreated . Cultures were grown for two more hours , upon which protein synthesis was stopped by addition of cyclohexamide . GFP fluorescence was measured by flow cytometry , using a Miltenyi MACSQuant VYB flow cytometer . The GFP signal was averaged for all duplicates and standard errors were calculated . All experiments were repeated at least twice ( total number of colonies analyzed: n≥4 ) and found to be in good agreement . Two tailed t-tests with unequal variances were performed to assess the statistical significance of the differences in GFP fluorescence values measured by flow cytometry for the different samples . All domain-rearrangement variants were tagged with GFP at their N-termini , as previously described [11] . Imaging was performed with an automated inverted Leica TCS SP8 confocal microscope . Mating assays were performed with minor modifications to a previously described method [11] . Specifically , each “a-type” individual deletion strain ( SO992 , W303-derived , trp1 , leu2 , ura3 , his3 , ADE2 can1 ) described above was transformed with appropriate plasmids encoding each domain-rearrangement variant to be tested . Equal amounts of “A-type” cells transformed with each variant ( or controls ) were mixed with WT “α-type” cells and deposited on the surface of a polycarbonate filter placed on a YPD plate and incubated for 3 hours at 30°C . Cells were then detached from the filters by vortexing and aliquots were plated on minimum synthetic media , or synthetic media lacking lysine . Mating efficiency was calculated as the number of colonies on minimum synthetic media divided by number of colonies on synthetic media lacking lysine [49] . Results were normalized according to the WT type strain . Averages from triplicates and standard errors were calculated . The experiments were repeated at least twice ( total number of colonies analyzed: n≥6 ) and found to be in good agreement . Site-direct mutagenesis was done by Quick Change , following the manufacturer's protocol ( Quick Change II Site-Directed Mutagenesis kit , Agilent ) . Mutations were verified by DNA sequencing . Proteins with domain compositions similar to those found in our experiments were identified in the Domain Club Database [10] . Hydrodynamic Radii for IDRs was calculated using the power law relation Rh = F * ρ0 * Nν [50] , where ρ0 is a constant that depends on persistence length , N is the number of residues in the polymer , ν is a scaling factor , and F is a correction factor that accounts for the net charge and Pro content of the IDR . The IDRs of Ste50 , Ste20 , and Ste11 were identified in the Pfam database . Specifically , we deleted the disordered regions between amino acids 156 and 230 in Ste50 , 408 and 578 in Ste20 , and between 258 and 354 in Ste11 .
|
Cells use complex protein interaction networks to sense and process external signals . Proteins involved in signaling are often composed of multiple functional units called domains . Because domains are modular , mutations that rearrange domains among proteins have the potential to result in the creation of novel proteins with altered functions . At an evolutionary timescale , domain rearrangements contribute to the functional diversification of signaling networks; at the shorter timescale of the life of an individual , domain rearrangements can impair cellular functions and lead to disease . Here , we investigated how domain-rearranging mutations alter the function of signaling networks , in particular when these mutations disrupt pre-existing proteins . We used as a model system the yeast mating signaling pathway , which shares many properties with more complex pathways active in human cells . Our results demonstrate that signaling networks are often robust to domain rearrangements that disrupt pre-existing genes . In addition , our experiments suggest a possible mechanism to explain this robustness: rather than being a rigid multi-protein machine , the yeast mating signaling complex is a dynamic ensemble with loose spatial constraints . Because of this , the changes in protein interaction partners caused by domain-rearrangement mutations can be accommodated without disrupting network function .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"biochemistry",
"synthetic",
"biology",
"biology",
"and",
"life",
"sciences",
"evolutionary",
"biology"
] |
2014
|
The Robustness of a Signaling Complex to Domain Rearrangements Facilitates Network Evolution
|
The Tax protein of human T-cell leukemia virus type 1 ( HTLV-1 ) is crucial for the development of adult T-cell leukemia ( ATL ) , a highly malignant CD4+ T cell neoplasm . Among the multiple aberrant Tax-induced effects on cellular processes , persistent activation of transcription factor NF-κB , which is activated only transiently upon physiological stimulation , is essential for leukemogenesis . We and others have shown that Tax induces activation of the IκB kinase ( IKK ) complex , which is a critical step in NF-κB activation , by generating Lys63-linked polyubiquitin chains . However , the molecular mechanism underlying Tax-induced IKK activation is controversial and not fully understood . Here , we demonstrate that Tax recruits linear ( Met1-linked ) ubiquitin chain assembly complex ( LUBAC ) to the IKK complex and that Tax fails to induce IKK activation in cells that lack LUBAC activity . Mass spectrometric analyses revealed that both Lys63-linked and Met1-linked polyubiquitin chains are associated with the IKK complex . Furthermore , treatment of the IKK-associated polyubiquitin chains with Met1-linked-chain-specific deubiquitinase ( OTULIN ) resulted in the reduction of high molecular weight polyubiquitin chains and the generation of short Lys63-linked ubiquitin chains , indicating that Tax can induce the generation of Lys63- and Met1-linked hybrid polyubiquitin chains . We also demonstrate that Tax induces formation of the active macromolecular IKK complex and that the blocking of Tax-induced polyubiquitin chain synthesis inhibited formation of the macromolecular complex . Taken together , these results lead us to propose a novel model in which the hybrid-chain-dependent oligomerization of the IKK complex triggered by Tax leads to trans-autophosphorylation-mediated IKK activation .
Human T-cell leukemia virus type 1 ( HTLV-1 ) is etiologically associated with adult T-cell leukemia ( ATL ) , an aggressive and lethal malignancy of CD4+ T cells , and with HTLV-1-associated myelopathy/tropical spastic paraparesis ( HAM/TSP ) [1 , 2] . The HTLV-1 provirus genome encodes a transactivator protein ( Tax ) , which is crucial for viral gene expression and the onset and development of ATL together with another viral protein , HBZ [3–5] . Tax aberrantly activates host cell transcription factors , including nuclear factor-κB ( NF-κB ) , cyclic AMP response element-binding protein ( CREB ) and serum responsive factor ( SRF ) , thereby perturbing transcriptional networks in host cells [6] . Among these factors , accumulating evidence indicates that persistent activation of NF-κB by Tax is crucial for T cell transformation and ATL development [7–9] . NF-κB plays critical roles in immune responses , inflammation , bone metabolism , cell proliferation and survival [10] . NF-κB is composed of five Rel/NF-κB family members—p50/p105 , p52/p100 , RelA , RelB and c-Rel—which form various combinations of homo- and heterodimers . NF-κB is sequestered in the cytoplasm with inhibitory proteins of the NF-κB family ( IκBs ) or NF-κB precursors . Two distinct pathways lead to activation of NF-κB . The canonical pathway is activated by cytokines , such as tumor necrosis factor ( TNF ) -α and interleukin ( IL ) -1 , whose stimulation leads to activation of the IκB kinase ( IKK ) complex , which is composed of the catalytic subunits IKKα and IKKβ and the regulatory subunit NEMO [11] . The IKK complex then induces phosphorylation and subsequent degradation of IκBα , which allows the p50/RelA heterodimer to translocate into the nucleus and activate target genes . In the noncanonical pathway , stimulation of CD40 , receptor activator of NF-κB ( RANK ) or lymphotoxin-β receptor results in the activation of IKKα in a NIK-dependent but IKKβ- and NEMO-independent manner . IKKα then phosphorylates the C-terminal ankyrin repeats of p100 , which forms heterodimers with RelB in the cytoplasm , leading to the proteasome-dependent selective degradation of the p100 C-terminal end to generate p52 [12] . The resulting p52/RelB heterodimer translocates into the nucleus and activates target genes . Tax is able to activate both canonical and noncanonical pathways , which are thought to be coordinately involved in leukemogenesis [13] . The importance of ubiquitination in the regulation of NF-κB activity is well established [14] . Ubiquitination is catalyzed by three enzymes in a stepwise fashion [15] . Ubiquitin-activating enzyme E1 forms a thioester linkage with ubiquitin , and the activated ubiquitin is then transferred to the E2 ubiquitin-conjugating enzyme . E2 acts as an escort for ubiquitin to the subsequent enzyme , E3 ligase , which binds to both E2 and the substrate and catalyzes the formation of an isopeptide bond between carboxylic acid at the C-terminal end of ubiquitin and the epsilon amine of the lysine residue in the substrate . After the addition of a single ubiquitin to the substrate , more ubiquitin can be repeatedly added to the previously conjugated molecule , thereby yielding a polyubiquitin chain . Ubiquitin itself contains seven lysines ( K6 , K11 , K27 , K29 , K33 , K48 and K63 ) , each of which can participate in the formation of the ubiquitin chain , allowing seven linkage types [16] . In addition , ubiquitin can also be attached to the N-terminus of the proximal ubiquitin to generate a linear ubiquitin chain or Met1-linked ubiquitin chain ( M1 chain ) [17] . In the TNFR signaling pathway , several types of polyubiquitin chains cooperatively regulate IKK activation . Upon TNF-α stimulation , TNF receptor-1 recruits the adaptor TRADD , TRAF2/5 and cIAPs . cIAPs have E3 ligase activity and conjugate Lys11- and Lys63-linked ubiquitin chains ( K11 and K63 chains ) to RIP1 [18–20] . These polyubiquitin chains conjugated to RIP1 act as a scaffold for the formation of an active signaling complex containing transforming growth factor-β-activated kinase ( TAK ) -1 , TAK1-binding ( TAB ) 2/3 and linear ubiquitin chain assembly complex ( LUBAC ) [21] . LUBAC is composed of HOIL-1L , HOIP and Sharpin [22 , 23] . HOIP is the catalytic subunit while HOIL-1L and Sharpin are also required for the enzymatic activity of this complex . LUBAC conjugates M1 chains to NEMO [24] , which may induce oligomer formation or a conformational change of NEMO to activate the IKK complex . Although previous studies have shown that Tax binds to NEMO and induces constitutive activation of the IKK complex in a K63-chain-dependent manner [25–28] , the involvement of other types of polyubiquitin chains in Tax-induced IKK activation is still controversial [29] . In this study , we show that Tax induces generation of hybrids of K63 and M1 chains by recruiting LUBAC to the IKK complex , leading to the formation of the active macromolecular IKK complex . Thus , we propose a previously unidentified mechanism by which K63 and M1 chains cooperate in Tax-induced IKK activation .
We previously established a cell-free assay to analyze Tax-induced IKK activation , in which the addition of recombinant Tax protein purified from E . coli into S-100 cytosolic extracts prepared from the Jurkat human T cell line , HEK293T cell line or mouse embryonic fibroblast ( MEF ) cells results in IKK activation [27] . To investigate which types of polyubiquitin linkages are required for Tax-induced IKK activation , we took advantage of a cell-free assay because the addition of dominant-negative ( DN ) ubiquitin mutants containing a single lysine-to-arginine substitution ( K6R , K11R , K27R , K29R , K33R , K48R and K63R ) or N-terminal HA-tagged ubiquitin results in linkage type-specific blockage of polyubiquitination . Immunoblots probed with anti-phospho-IKKα/β and phospho-IκBα antibodies revealed that the addition of K27R , K63R or HA-ubiquitin inhibited Tax-induced IKK activation ( Fig 1A ) , suggesting that K27 , K63 and M1 chains are required for IKK activation by Tax . Addition of K11R or K33R ubiquitin reproducibly enhanced Tax-induced IKK activation , probably because their addition could enhance the generation of K27 , K63 or M1 chains . Note that phosphorylated IκBα is not degraded by proteasomes in a cell-free assay ( S1 Fig ) , although the amount of IκBα was slightly reduced concomitantly with IκBα phosphorylation in some experiments in this paper . This could be due to the manufacturer-noted preferential binding of the anti-IκBα antibody used for immunoblotting to the non-phosphorylated form of IκBα . To identify the E2 ubiquitin-conjugating enzymes involved in Tax-induced IKK activation , a cell-free assay was performed using cytosolic extracts prepared from HEK293T cells expressing a series of E2 DN mutants , in which an active Cys residue was substituted with Ala . Expression of the Ubc13 DN mutant almost completely inhibited IKK activation , whereas other E2 DN mutants did not ( Fig 1B ) . A cell-free assay using the extract from Ubc13−/− MEFs further confirmed that Ubc13 is required for Tax-induced IKK activation ( Fig 1C ) , which is consistent with previous reports based on experiments using intact Ubc13−/− MEFs and a cell-free assay using the extract from Ubc13-knockdown cells [28 , 30] . Low-level Tax-induced phosphorylation of IκBα was observed in Ubc13−/− MEFs , which could be due to residual Ubc13 attributable to incomplete gene disruption by the Cre/loxP system . The lack of candidate E2 enzymes other than Ubc13 suggests that several E2 enzymes may redundantly generate K27 and M1 chains or that E2 enzymes not tested here could be involved . Because it has been reported that RNF8 , as an E3 ubiquitin ligase , is partially involved in the Tax-induced generation of K63 chains [28] , we checked whether other E3 enzymes capable of generating K63 chains are involved [18 , 31–33] . Cytosolic extracts were prepared from cIAP1/cIAP2-deficient ( Birc2−/−/Birc3−/− ) , TRAF2/TRAF5-deficient ( Traf2−/−/Traf5−/− ) , TRIM25-deficient ( Trim25−/− ) and Riplet-deficient ( Rnf135−/− ) MEFs [33–36] and were subjected to a cell-free assay . None of the extracts derived from the mutant cells showed reduced IKK activation ( S2 Fig ) , suggesting that these E3 enzymes are not involved in Tax-induced IKK activation . In addition , TRAF6 , another E3 enzyme , has been shown to be dispensable in Tax-induced IKK activation but instead can work together with Ubc13 to generate K63 chains for cytokine-induced IKK activation [27 , 37] . We then hypothesized that Tax itself may be an E3 ligase as recently proposed [29] , since Tax contains a putative zinc finger domain at its N-terminus ( S3A Fig ) [38] and the zinc finger domain may act as a catalytic domain of E3 ligase as previously shown in the zinc finger of A20 [39] . Some zinc finger mutants of Tax failed to activate the IKK complex and NF-κB ( S3B and S3C Fig ) , indicating that the zinc finger of Tax is crucial for IKK activation . Although recombinant Tax purified from either E . coli or Sf9 cells can efficiently activate IKK ( S3D Fig ) , neither of them induced polyubiquitination in the presence of E2 enzymes including UbcH5c , UbcH7 and Ubc13/Uev1A under conditions that allow TRAF6 to generate polyubiquitin chains together with Ubc13/Uev1A ( S3E Fig ) . These results strongly suggest that Tax itself does not possess E3 ligase activity . To further confirm the requirement for M1 chains , cytosolic extracts derived from MEFs that lack each component of LUBAC ( the only known E3 ligase complex that catalyzes M1 chain generation ) were tested . Tax failed to induce IKK activation when cytosolic extracts from HOIL-1L-deficient ( Rbck1−/− ) MEFs , Sharpin-deficient ( cpdm ) MEFs or MEFs in which the RING-IBR-RING region ( the catalytic center ) of HOIP was ablated ( HOIPΔlinear ) were used ( Fig 2A–2C ) [40] . To confirm the requirement for LUBAC for Tax-induced IKK activation in intact cells , Sharpin-deficient MEFs were infected with a Tax-expressing retrovirus , and subsequent phosphorylation of IKK and IκBα was detected by immunoblotting . Tax-induced IKK activation was significantly reduced in cells that lack LUBAC activity ( Fig 2D ) . Taken together , these results clearly indicate that LUBAC is crucial for Tax-induced IKK activation . To understand how LUBAC is involved in Tax-induced IKK activation , we first investigated whether LUBAC binds to Tax . When Tax was immunoprecipitated with an anti-Tax antibody after incubation with Jurkat cytosolic extracts , HOIP and Sharpin were co-immunoprecipitated with Tax ( Fig 3A ) , indicating that Tax interacts with LUBAC . In addition , the Tax mutant M22 , which is incapable of activating NF-κB due to a lack of binding ability to NEMO [30] , also bound to HOIP and Sharpin ( Fig 3A ) , indicating that the binding of Tax to LUBAC is not mediated by the IKK complex . This result led us to hypothesize that Tax acts as an adaptor in the formation of a multi-protein complex composed of LUBAC , Tax and the IKK complex . To test this hypothesis , the IKK complex was immunoprecipitated with an anti-Flag antibody from cytosolic extracts of Jurkat cells expressing Flag-NEMO . HOIP and Sharpin were recruited to the IKK complex in the presence of Tax , but M22 failed to recruit LUBAC to the IKK complex ( Fig 3B ) , indicating that Tax functions as an adaptor to recruit LUBAC to the IKK complex . We then sought to determine whether Tax also acts as a bridge between LUBAC and the IKK complex in an intact Jurkat human T cell line . JPX-9 , a Jurkat-derived cell line in which Tax expression is induced by Cd2+ treatment [41] , was first cultured in the presence or absence of Cd2+ , and cell lysates were subjected to immunoprecipitation using an anti-Tax antibody . HOIP and Sharpin were included in the immunoprecipitates only when Tax was induced ( Fig 3C ) . These results indicate that Tax associates with LUBAC in intact T cells . Interestingly , the slower-migrating form of HOIP was observed only when the IKK complex was activated by Tax ( Fig 3A ) . This band shift was due to the phosphorylation of HOIP because Phos-tag SDS-PAGE analysis identified slower-migrating bands ( S4A Fig ) . Treatment of lysates with the IKKβ inhibitor TPCA-1 resulted in the disappearance of the slower-migrating bands in a dose-dependent manner ( S4B Fig ) , suggesting that IKKβ phosphorylates HOIP during Tax-induced IKK activation . The significance of HOIP phosphorylation in IKK activation remains to be elucidated . Given that HOIP binds to K63 chains but not M1 chains [42] , we hypothesized that K63 chains are required for the binding of LUBAC to the IKK complex . To test this possibility , we first investigated whether the addition of DN ubiquitin mutants would inhibit the Tax-mediated binding of LUBAC to the IKK complex . The addition of K63R or HA-tagged ubiquitin inhibited Tax-induced IKK activation ( Fig 1A ) , whereas the Tax-mediated binding of LUBAC to the IKK complex was not affected ( Fig 3D ) . These results indicate that K63 and M1 chains are not required for the binding of Tax to LUBAC and the IKK complex . To determine which components of LUBAC bind to Tax and also whether the binding is direct , an in vitro binding assay was performed using purified recombinant proteins . Purified GST-HOIL-1L , GST-Sharpin or GST-HOIP was incubated with His6-Tax and subjected to GST pull-down assay . GST-HOIL-1L and GST-HOIP bound to His6-Tax , whereas GST-Sharpin did not ( Fig 3E ) , indicating that HOIL-1L and HOIP directly bind to Tax . To elucidate the molecular basis of the binding of HOIL-1L or HOIP to Tax , a series of deletion mutants of HOIL-1L and those of HOIP were tested by co-immunoprecipitation assay . HOIL-1L ΔUBL and HOIP ΔRBR failed to bind to Tax , whereas the other mutants of HOIL-1L and HOIP proteins bound to Tax as efficiently as the full-length protein ( Fig 3F and 3G ) . These results indicate that HOIL-1L and HOIP interact directly with Tax through their UBL and RBR domains , respectively . To determine how the Tax-induced generation of polyubiquitin chains leads to IKK activation , Jurkat cytosolic extracts were incubated in the absence or presence of recombinant Tax , and the reaction mixtures were then subjected to immunoprecipitation with an anti-NEMO antibody . The resulting immunoprecipitates were immunoblotted with either an anti-ubiquitin ( Ub ) antibody that can recognize monoubiquitin and any type of polyubiquitin linkages or an anti-M1 chain-specific antibody . Both antibodies clearly detected smeared bands only when cytosolic extracts were incubated with Tax ( Fig 4A , lane 2 ) , indicating that M1 chains were associated with the IKK complex in a Tax-dependent manner . To further characterize the IKK complex-associated ubiquitin chains , the immunocomplexes precipitated with an anti-NEMO antibody were then treated with the following chain type-specific deubiquitinases ( DUBs ) : Otubain-1 for K48 chains [43 , 44] , associated molecule with the SH3 domain of STAM ( AMSH ) for K63 chains [42 , 45] , OTULIN for M1 chains [46] , and ubiquitin-specific protease 2 ( USP2 ) for any type of polyubiquitin chain [47] . Otubain-1 treatment did not affect the smears detected by the anti-Ub or anti-M1 chain antibody ( Fig 4A , lane 3 upper and lower ) , indicating that K48 chains were nearly nonexistent in the complex . In contrast , AMSH treatment almost completely abolished the smears detected by the anti-Ub antibody ( Fig 4A , lane 4 upper ) , and OTULIN treatment completely abolished the smears detected by the anti-M1 chain antibody ( Fig 4A , lane 5 lower ) . These results indicated that both K63 and M1 chains were associated with the IKK complex . These results were further supported by the quantification of different ubiquitin chain types associated with the IKK complex via the ubiquitin-AQUA method using mass spectrometry [48] . While residual amounts of K48 chains were detected irrespective of the presence of Tax , an approximately 2:1 ratio of K63 to M1 chain linkages was significantly associated with the IKK complex only when cytosolic extracts were incubated with Tax ( Fig 4B ) . Interestingly , when an immunoblot of the AMSH-treated IKK complex was probed with the anti-M1 chain antibody , extremely high-molecular-weight ubiquitin-containing complexes ( EHUCs ) , which remained at the top of separating gels ( arrows in Fig 4A and 4C ) , were degraded ( Fig 4A , lane 4 lower ) , indicating that EHUCs are polyubiquitin chains that include both K63 and M1 linkages in a single chain . This notion was also supported by experiments showing that Tax failed to generate EHUCs when either the generation of K63 chains or that of M1 chains was blocked ( Fig 4C , lanes 5 and 6 ) . Furthermore , when an immunoblot of the OTULIN-treated IKK complex was probed with the anti-Ub antibody , the abundance of EHUCs was found to be significantly reduced while ladders between 17 and 75 kDa appeared ( Fig 4A , lanes 5 and 7 upper , note that the same sample was applied in lanes 5 and 7 ) . Importantly , among these ladders , four bands ( dots in Fig 4A , lanes 5 and 7 upper ) migrated almost identically to the bands corresponding to trimer ( Ub3 ) , tetramer ( Ub4 ) , pentamer ( Ub5 ) and hexamer ( Ub6 ) of recombinant K63 chains ( arrowheads in Fig 4A , lane 8 upper ) . These results clearly indicate that K63/M1-linked hybrid chains are associated with the IKK complex activated by Tax . To address whether these IKK complex-associated polyubiquitin chains are required for Tax-induced IKK activation , cell-free assays were performed in the presence of various DUBs . AMSH , OTULIN and USP2 , but not Otubain-1 , inhibited the phosphorylation of IκBα induced by Tax ( Fig 4D ) , indicating that generation of both K63 and M1 chains is essential for Tax-induced IKK activation . Therefore , generation of IKK complex-associated K63/M1-linked hybrid chains is likely to be essential for Tax-induced IKK activation . In the cytokine-induced NF-κB signaling pathway , IKK activation requires the formation of unanchored K63 chains or NEMO-conjugated M1 chains [24 , 49] . To understand the roles of unanchored and substrate-conjugated ( anchored ) chains in Tax-induced IKK activation , cell-free assays were performed in the presence of Isopeptidase T ( IsoT ) , a DUB specific for unanchored chains , or the OTU domain of the L protein of Crimean Congo hemorrhagic fever virus ( viral OTU ) , a DUB specific for substrate-anchored chains . IsoT inhibited the Tax-induced phosphorylation of IKK and IκBα but not their polyubiquitination-independent phosphorylation by MEKK1 ( Fig 5A ) . In addition , viral OTU , but not its catalytic inactive mutant ( 1A ) , inhibited the Tax-induced phosphorylation of IKK and IκBα ( Fig 5B ) . These results indicate that both unanchored and substrate-anchored polyubiquitin chains are required for Tax-induced IKK activation . Several studies have shown that ubiquitination of Tax is required for IKK activation [50–52] . Among ten lysine residues present in Tax , ubiquitination of the C-terminal seven lysines ( K4 to K10 ) are required for Tax-induced IKK activation in intact cells [50] . To examine whether Tax requires similar ubiquitination for IKK activation in our cell-free system , recombinant Tax mutants containing lysine-to-arginine mutations at the three N-terminal lysines ( K1-3R ) or at the seven C-terminal lysines ( K4-10R ) were generated . Both Tax-WT and the K1-3R mutant induced IKK activation equally well , whereas the K4-10R mutant did not ( Fig 5C left ) . In addition , the K4-10R mutation significantly reduced Tax ubiquitination ( Fig 5C right ) . These results strongly suggest that the polyubiquitin chains conjugated to Tax belong to the class of substrate-anchored polyubiquitin chains required for Tax-induced IKK activation . Because LUBAC induces polyubiquitination at K285 and K309 of NEMO in cytokine-induced IKK activation [24] , we next investigated whether these lysine residues are required for Tax-induced IKK activation . Human NEMO-WT or its mutant ( K285R/K309R ) was introduced into NEMO-deficient ( Ikbkg−/− ) MEFs , and cell-free assays were performed . Tax induced phosphorylation of IKK and IκBα equally well in NEMO-WT- and NEMO ( K285R/K309R ) -expressing cytosolic extracts ( Fig 5D ) , indicating that polyubiquitination at K285 and K309 of NEMO is dispensable for Tax-induced IKK activation . It has been reported that the binding of NEMO to K63 and M1 chains is required for cytokine-induced IKK activation . NEMO binds to K63 chains through the C-terminal NZF domain and to M1 chains through the UBAN domain [53 , 54] . To determine the requirement for the binding ability of NEMO to K63 or M1 chains , cytosolic extracts were prepared from NEMO-deficient MEFs expressing mouse NEMO-WT , its mutant ( R309A/R312A/E313A ) lacking the ability to bind to M1 chains or another NEMO mutant ( H406A/C410A ) lacking the ability to bind to K63 chains . Tax failed to induce IKK activation in cytosolic extracts expressing the NEMO ( R309A/R312A/E313A ) or NEMO ( H406A/C410A ) mutant ( Fig 5E ) , indicating that the binding ability of NEMO to both K63 and M1 chains is required for Tax-induced IKK activation . Because the ability of NEMO to bind to both K63 and M1 chains and the Tax-induced generation of the IKK complex-associated K63/M1-linked hybrid chains are required for the activation of IKK by Tax , we hypothesized that the macromolecular complex of IKK may be formed through multivalent interactions between polyubiquitin chains and NEMO , facilitating trans-autophosphorylation between the IKK complexes , thereby inducing IKK activation . To test this possibility , we performed cell-free assays in the presence or absence of DN ubiquitin mutants , and the reaction mixtures were subjected to Blue native-PAGE , which can be used to determine the size and composition of native protein complexes [55] , followed by immunoblotting . Probing the immunoblots with an anti-NEMO antibody revealed that the NEMO-containing macromolecular complex ( arrowheads in Fig 5F left ) was formed only in the presence of Tax in addition to the regular complex of approximately 600 kDa ( dots in Fig 5F ) , which is observed as an inactive IKK complex in the absence of Tax . Interestingly , the formation of the macromolecular complex was abrogated when the extract was incubated with DN ubiquitin mutants ( K27R , K63R , or HA-Ub ) that also inhibit Tax-induced IKK activation ( Fig 5F left ) . Furthermore , probing the immunoblots with an anti-p-IKKα/β antibody revealed that activated IKK was observed only when the macromolecular complex was formed and that activated IKK was included in the macromolecular complex ( Fig 5F right ) . These results strongly suggest that polyubiquitination-dependent formation of the macromolecular IKK complex triggers IKK activation . To investigate the physiological significance of LUBAC in HTLV-1-infected cells , we first checked the interaction between Tax and LUBAC . Lysates prepared from the HTLV-1-infected human T cell line HUT102 , which is derived from a mycosis fungoides patient [56] , were subjected to immunoprecipitation using an anti-Tax or a control antibody . HOIP and Sharpin were precipitated only when Tax was precipitated by the anti-Tax antibody ( Fig 6A ) . To further confirm the physiological interaction between Tax and LUBAC , MT-2 or MT-4 cell lines , T cell lines transformed by co-culture with HTLV-1-producing ATL cells [57 , 58] , were used . HOIP and Sharpin were co-precipitated with Tax using an anti-Tax antibody . However , neither HOIP nor Sharpin were precipitated when Tax was knockdown ( Fig 6B ) . These results indicate that Tax interacts with endogenous LUBAC in three distinct HTLV-1-infected cell lines . To understand whether LUBAC is involved in IKK activation in HTLV-1-infected cells , effect of HOIP knockdown on IKK activation was analyzed . Immunoblotting with an anti-pIKKα/β antibody revealed that HOIP knockdown blocked IKK activation in MT-2 and MT-4 cells ( Fig 6C ) . Consistently , expression levels of the NF-κB target genes were notably reduced in HOIP-knockdown MT-4 cells ( Fig 6D ) . Moreover , HOIP knockdown significantly suppressed cell proliferation ( Fig 6E ) . Taken together , these results show that LUBAC-mediated M1 chain formation is required for NF-κB activation leading to target gene expression and cell proliferation in HTLV-1-infected cells .
Extensive studies on the role of Tax in ATL development have demonstrated that Tax is involved in leukemogenesis largely through its ability to constitutively activate NF-κB [7–9] . It has been known for almost two decades that the binding of Tax to NEMO is required for IKK activation , but the precise molecular mechanisms by which this binding leads to IKK activation remain to be elucidated . We and other groups have shown that the Tax-induced generation of K63 chains is crucial for IKK activation [27 , 28 , 30] . Furthermore , Ho et al . [28] recently provided clear evidence that RNF8 acts as an E3 ligase to generate K63 chains for IKK activation . As an extension of these previous results , we propose a novel molecular model for Tax-induced IKK activation in which LUBAC , together with unidentified E3 ligases for K63 chains , generate K63/M1-linked hybrid chains to form the active macromolecular Taxisome , composed of LUBAC , Tax and the active IKK complex , thereby establishing persistent NF-κB activation ( Fig 7 ) . Cell-free experiments allowed us to address the effects of chain type-specific blocking of polyubiquitin synthesis on critical steps of Tax-induced IKK activation . Several lines of evidence presented here support our model . 1 ) In vitro binding and immunoprecipitation experiments revealed that Tax can bind to both the IKK complex and LUBAC to form an inactive pre-Taxisome without the generation of polyubiquitin chains , whereas activation of the IKK complex by Tax requires the synthesis of the K27 , K63 and M1 chains . 2 ) Genetic evidence revealed that both Ubc13 ( an E2 enzyme for K63 chain synthesis ) and each component of LUBAC ( the only E3 enzyme for M1 chain ) are crucial for Tax-induced IKK activation . 3 ) Mass spectrometric analyses revealed that both K63 and M1 chains are associated with the IKK complex only in the presence of Tax . 4 ) Cell-free experiments with chain type-specific DUBs revealed that K63/M1-linked hybrid chains are associated with the active IKK complex . 5 ) Tax-induced IKK activation requires the ability of NEMO to bind to both K63 and M1 chains . 6 ) The formation of the active macromolecular IKK complex ( active Taxisome ) requires the synthesis of the K27 , K63 and M1 chains . Based on 4 ) , 5 ) and 6 ) , a single hybrid chain may bind to multiple NEMO molecules , and a single NEMO may act as a bridge between the hybrid chains , which may explain why generation of the hybrid chain is required for the macromolecular IKK complex . Regarding how the formation of the macromolecular complex leads to IKK activation , there are two potential mechanisms . The first is that the formation of the macromolecular complex could induce an efficient physical interaction between IKK complexes , so that trans-autophosphorylation between IKK complexes results in full activation of IKK . The second possible mechanism is that the formation of the macromolecular complex could somehow recruit an IKK kinase ( IKKK ) such as TAK1 , which phosphorylates and activates IKK in response to cytokine stimulation [59] . We previously reported that MAP3Ks , including MEKK1 , MEKK3 , NIK , TPL-2 and TAK1 , are dispensable for Tax-induced IKK activation [60] . In addition , our extensive proteomics analysis of the Tax-activated IKK complex failed to identify any IKKK candidates [61] , leading us to prefer the trans-autophosphorylation model . In contrast to our model , conflicting results have been reported on the following three points . The first point is the involvement of IKKK . Yin et al . [62] demonstrated that a dominant negative MEKK1 mutant inhibits IKK activation induced by Tax . Ho et al . [28] and Wu et al . [63] reported that Tax fails to activate the IKK complex in TAK1-knockdown HeLa cells or TAK1-knockout MEFs . This discrepancy may due to the experimental conditions including cell types used . Extensive analysis of the macromolecular complex and genetic and biochemical experiments in T cells are required to determine the involvement of any IKKK in Tax-induced IKK activation . The second point concerns the E3 ligase for K63 chain formation . Wang et al . [29] recently reported that Tax acts as an E3 ubiquitin ligase for IKK activation through synthesis of mixed-linkage polyubiquitin chains and that K63 chains are dispensable for Tax-induced IKK activation . However , we show that K63 chains are essential and could not demonstrate the E3 activity of Tax under any conditions we tested . Further identification of factors involved in Tax-induced IKK activation may explain these discrepancies . The third point concerns the subcellular localization of the active Taxisome . Pujari et al . [64] reported that membrane-associated Cell adhesion molecule 1 ( CADM1 ) functions as a scaffold for Tax and Ubc13 to activate the IKK complex in intact cells . Since the S-100 fraction used in the cell-free system does not contain membrane fractions , we believe that Tax can interact with Ubc13 in the absence of CADM1 . However , our data do not exclude the possibility that , in the cell free assay , Tax induces IKK activation without molecules required for subcellular localization in intact T cells . Precise comparison of IKK activation in the cell-free system with that in intact cells will explain regulation of Taxisome formation in HTLV-1 infected cells Inconsistent with our model , the regular complex ( dots in Fig 5F ) , which was inactive in the absence of Tax , appeared to be activated in the presence of Tax in addition to the macromolecular complex . This could occur because the macromolecular complex is unstable , such that the activated macromolecular complex dissociates into the active regular-size complex , or because the active macromolecular complex is able to transiently associate with and activate the regular complex . However , we cannot completely rule out the possibility that the IKK complex can be activated without the formation of the macromolecular complex in the presence of Tax . Interestingly , cell-free experiments using IsoT and viral OTU revealed that both unanchored and anchored polyubiquitin chains are required for Tax-induced IKK activation . Although the critical roles of unanchored and anchored polyubiquitin chains in Tax-induced IKK activation are still controversial , our experiments using Tax KR mutants support the idea that the polyubiquitin chains conjugated to Tax belong to the class of substrate-anchored polyubiquitin chains required for Tax-induced IKK activation . By taking advantage of specific DUBs that degrade only one type , we showed for the first time that unanchored and anchored polyubiquitin chains cooperate in Tax-induced IKK activation . Although the addition of the DN ubiquitin mutant K27R inhibited IKK activation by Tax , we did not detect K27 chains associated with the active Taxisome via the ubiquitin-AQUA method . This may be because an undetectable amount of K27 chains is involved in the generation of the hybrid chains as their component or because K27 chains act as initial triggers of the hybrid chains synthesis but are dissociated from the Taxisome afterwards . Further studies are needed to identify the precise role of K27 chains in Tax-induced IKK activation . Tax binds to LUBAC through the UBL domain of HOIL-1L and RBR domain of HOIP . The UBL domain of HOIL-1L is involved in the association with HOIP , while the RBR domain of HOIP is the catalytic active site [66] . Therefore , the association of Tax with LUBAC may activate the E3 activity of LUBAC upon HTLV-1 infection , which may be one of the critical events in the onset of ATL . In accord with this scenario , small compounds that specifically block the interaction between Tax and HOIL-1L or between Tax and HOIP could be used as novel therapeutic approaches for ATL and HAM .
Human cDNAs encoding dominant-negative mutants of E2 enzymes were generated via PCR and inserted into the pRK5 vector . pMRX-Cre was obtained from S . Akira ( Osaka University ) . Expression vectors for HOIL-1L-HA , Myc-HOIP and their deletion mutants were generated as previously described [66] . Human and mouse cDNAs encoding NEMO and NEMO mutants were inserted into the retrovirus vector pMXs obtained from T . Kitamura ( University of Tokyo ) . Viral OTU cDNA was obtained from A . García-Sastre ( Icahn School of Medicine at Mount Sinai ) . 3xκB-luc was obtained from S . Miyamoto ( University of Wisconsin-Madison ) . Tax cDNA was obtained from J . Fujisawa ( Kansai Medical University ) . Tax mutants were generated via PCR and inserted into the pCG vector . The following antibodies were used: anti-p-IκBα ( 9246 ) , anti-IκBα ( 9242 ) , anti-p-IKKα/β ( 2697 ) , anti-NEMO ( 2695 ) and anti-ubiquitin ( 3936 ) ( Cell Signaling Technology ) ; anti-GST ( sc-459 ) , anti-Myc ( sc-789 ) and anti-HA probe ( sc-805 ) ( Santa Cruz Biotechnology ) ; anti-Flag M2 ( F3165 ) ( Sigma ) ; anti-tubulin ( CP06 ) ( Calbiochem ) ; anti-HOIL-1L ( NBP-1-88301 ) ( Novus Biologicals ) ; anti-HOIP ( ARP43241 ) ( Aviva Systems Biology ) ; anti-Sharpin ( 14626-1-AP ) ( Proteintech ) ; anti-linear polyubiquitin-specific ( AB130 ) ( LifeSensors ) ; and anti-Ubc13 ( 37–1100 ) ( ThermoFisher Scientific ) . The anti-Tax antibody was generated as previously described [67] . HEK293T cells ( purchased from ATCC ) , Plat-E cells ( provided by T . Kitamura ) and mouse embryonic fibroblasts ( provided by J . Silke , The Walter and Eliza Hall Institute , or established by us ) were maintained in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) . Jurkat cells ( purchased from ATCC ) , the Jurkat-derived cell line JPX-9 cells ( provided by K . Ohtani , Kwansei Gakuin University ) , and HTLV-1-infected cell line HUT102 , MT-2 , and MT-4 cells ( provided by J . Fujisawa ) were maintained in RPMI1640 supplemented with 10% heat-inactivated FBS . Sf9 cells were maintained in Sf900IIISFM ( Thermo Fisher Scientific ) supplemented with 10% FBS . The transfection of HEK293T cells was performed by the calcium phosphate method . siRNAs were transfected using NEPA21 Super Electroporator ( NEPAGENE ) . Control stealth siRNA and the following stealth siRNAs ( Thermo Fisher Scientific ) were used: HOIP-1 sense/anti-sense , 5′-GGUACUGGCGUGGUGUCAAGUUUAA-3′/5′-UUAAACUUGACACCACGCCAGUACC-3′; HOIP-2 sense/anti-sense , 5′-CACCACCCUCGAGACUGCCUCUUCU-3′/5′-AGAAGAGGCAGUCUCGAGGGUGGUG-3′ . For lentivirus production , HEK293T cells were transfected with the self-inactivating lentiviral vector construct , the packaging construct and the VSV-G- and Rev-expressing construct . After 48 h of incubation , culture supernatants were collected and centrifuged at 50 , 000 x g for 1 h at 20°C to concentrate lentivirus . MT-2 or MT-4 cells were infected with the lentivirus at 400 x g for 2 h at 20°C . After 48 h , puromycin ( Wako ) was added to the medium , and puromycin-resistant cell pools were used for further experiments . The following target sequences were used: Tax-1 , 5′-GGCCTTCCTCACCAATGTTCC-3′; Tax-2 , 5′-GGCAGATGACAATGACCATGA-3′; Tax-3 , 5′-GCCTACATCGTCACGCCCTAC-3′; HOIP-1 , 5′-GCTGCAGCTTTCAGAATTTGA-3′; HOIP-2 , 5′-GCACTGCCCATCCTGTAAACA-3′; HOIP-3 , 5′-GCTCCTTTGGCTTCATATATG-3′; Control , 5′-GATTTCGAGTCGTCTTAATGT-3′ . For retrovirus production , Plat-E cells were transfected with pMRX-Cre vector . After 24 h , culture supernatants were collected , and Ubc13+/+ or Ubc13fl/fl MEFs were incubated with the retrovirus containing polybrene ( 10 μg/ml; Sigma-Aldrich ) for 8 h . After 48 h , puromycin was added to the medium , and puromycin-resistant cell pools were used for further experiments . His6-Tax , His6-M22 , GST-viral OTU ( WT ) and its catalytic inactive mutant 1A were expressed in E . coli and purified . His6-TRAF6 and His6-Tax were expressed in Sf9 cells using the Bac-to-Bac Baculovirus Expression System ( Thermo Fisher Scientific ) and purified . GST , GST-HOIL-1L , GST-HOIP and GST-Sharpin were generated using the wheat germ cell-free protein synthesis system and purified [68] . Ubiquitin ( U-100H ) , ubiquitin mutants ( K6R ( UM-K6R ) , K11R ( UM-K11R ) , K27R ( UM-K27R ) , K29R ( UM-K29R ) , K33R ( UM-K33R ) , K48R ( UM-K48R ) and K63R ( UM-K63R ) ) , HA-ubiquitin ( U-110 ) , IsoT ( E-320 ) , Otubain-1 ( E-522B ) , AMSH ( E-548B ) , OTULIN ( E-558 ) , USP2 ( E-504 ) , His6-UBE1 ( E-304 ) , UbcH5c ( E2-627 ) , UbcH7 ( E2-640 ) and His6-Ubc13/Uev1A ( E2-664 ) were purchased from BostonBiochem . Recombinant K48- and K63-linked Ub2-Ub7 chains ( UC-230 , UC-330 ) were purchased from BostonBiochem . Recombinant M1-linked Ub2-Ub7 chains ( BML-UW1010-0100 ) were purchased from Enzo Life Sciences . Jurkat cells and MEFs were suspended in hypotonic buffer ( 10 mM Tris-HCl ( pH 7 . 5 ) , 1 . 5 mM MgCl2 , 10 mM KCl , 0 . 5 mM dithiothreitol ( DTT ) and protease inhibitor cocktail ( Roche ) ) and then lysed with a Dounce homogenizer . Cell debris was removed via ultracentrifugation at 100 , 000 x g for 1 h at 4°C to prepare the S-100 cytosolic extract . Cytosolic extracts ( 10 mg/ml ) were incubated with recombinant His6-Tax in ATP buffer ( 50 mM Tris-HCl ( pH 7 . 5 ) , 5 mM MgCl2 , 2 mM ATP , 5 mM NaF , 20 mM β-glycerophosphate , 1 mM Na3VO4 , and protease inhibitor cocktail ) in the presence or absence of various recombinant DUBs . After incubation at 30°C for 1 h , the reaction mixtures were subjected to immunoblotting or immunoprecipitation . Cells were lysed in IP buffer ( 20 mM Tris-HCl ( pH 7 . 5 ) , 150 mM NaCl , 2 mM EDTA , 1 mM MgCl2 , 10 mM NaF , 1% NP-40 , 10 mM β-glycerophosphate , 1 mM Na3VO4 , 1 mM DTT , 5 mM N-ethylmaleimide ( NEM ) and protease inhibitor cocktail ) , followed by centrifugation at 22 , 000 x g for 15 min at 4°C to remove the insoluble fraction . For detection of polyubiquitination of Tax , the reaction mixtures were boiled for 10 min in the presence of 1% SDS to remove noncovalently attached proteins . The mixtures were then diluted 10-fold in IP buffer to reduce the SDS concentration to 0 . 1% . The cell lysates or the cell-free reaction mixtures were subsequently incubated with the antibodies plus protein G-sepharose . The immunoprecipitates were washed five times and subjected to immunoblotting . For immunoblotting , immunoprecipitates or cell lysates were separated via SDS-PAGE and transferred to PVDF membranes ( Immobilon P , Millipore ) . The membranes were then incubated with the primary antibodies . Immunoreactive proteins were visualized with anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase , followed by processing with an ECL detection system . Glutathione sepharose was incubated with 300 ng of GST , GST-HOIL-1L , GST-HOIP or GST-Sharpin at 4°C for 1 h in binding buffer ( 50 mM Tris-HCl ( pH 7 . 5 ) , 150 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 1 mM DTT , 2 . 5 mg/ml BSA and protease inhibitor cocktail ) . The beads were then incubated with 500 ng of His6-Tax at 4°C for 1 h . After incubation , the beads were washed and subjected to immunoblotting . Jurkat cytosolic extracts ( 10 mg/ml ) were incubated with recombinant His6-Tax in ATP buffer . After incubation at 30°C for 1 h , the reaction mixtures were subjected to immunoprecipitation with an anti-NEMO antibody in IP buffer . The immunoprecipitates were washed three times with IP buffer without NEM and incubated with Otubain-1 ( 5 μM ) , AMSH ( 5 μM ) , OTULIN ( 5 μM ) or USP2 ( 5 μM ) at 37°C for 1 h in DUB buffer ( 50 mM HEPES-KOH ( pH 7 . 5 ) , 100 mM NaCl , 1 mM MnCl2 , 0 . 01% Brij-35 and 2 mM DTT ) . After incubation , the reaction mixtures were subjected to immunoblotting . Jurkat cytosolic extracts ( 10 mg/ml ) were incubated with recombinant His6-Tax in ATP buffer . After incubation at 30°C for 1 h , the reaction mixtures were subjected to immunoprecipitation with an anti-NEMO antibody in IP buffer . The immunoprecipitates were analyzed via the ubiquitin-AQUA method as described previously [48] . The immunoprecipitates were separated through SDS-PAGE , and the gel region above 50 kDa was subjected to in-gel trypsinization . The extracted peptides were analyzed with a Q Exactive mass spectrometer in targeted MS/MS mode together with 10 fmol of ubiquitin AQUA peptides . After the cell-free reaction , the reaction mixtures were mixed with NativePAGE Sample Buffer ( Thermo Fisher Scientific ) . Electrophoresis was performed using NativePAGE Running Buffer ( Thermo Fisher Scientific ) containing 0 . 002% G-250 . The gels were soaked in denaturation buffer ( 10 mM Tris-HCl ( pH 6 . 8 ) , 1% SDS and 0 . 006% 2-mercaptoethanol ) for 30 min at 60°C , followed by transfer to PVDF membranes and immunoblotting . The total RNA was isolated from MT-4 cells transfected with control or HOIP siRNA with Trizol reagent ( Thermo Fisher Scientific ) . cDNA was synthesized from 2 . 0 μg of total RNA with Prime ScriptII ( Takara ) . Quantitative real-time PCR analysis was performed on CFX Connect ( Bio-Rad ) . The level of GAPDH expression was used to normalize the data . The following primers were used: IL-6 sense/anti-sense , 5′-CCTGAACCTTCCAAAGATGGC-3′/5′-TTCACCAGGCAAGTCTCCTCA-3′; IL-1β sense/anti-sense , 5′-TTCGACACATGGGATAACGAGG-3′/5′-TTTTTGCTGTGAGTCCCGGAG-3′; MMP-9 sense/anti-sense , 5′-ATGTACCGCTTCACTGAGGG-3′/5′-TCAGGGCGAGGACCATAGAG-3′; GAPDH sense/anti-sense , 5′-TGCACCACCAACTGCTTAGC-3′/5′-GGCATGGACTGTGGTCATGAG-3′ . HEK293T cells were transfected with the plasmids encoding wild type or various Tax mutants together with 20 ng of luciferase reporter ( 3xκB-luc ) and 30 ng of β-actin-β-galactosidase plasmid . After 48 h , the luciferase activity was measured using the Luciferase Assay System ( Toyo Ink ) . β-galactosidase activity was used to normalize the transfection efficiency . The reactions were performed at 37°C for 1 h in the reaction buffer containing His6-UBE1 ( 0 . 1 μM ) , the indicated E2 ( 0 . 2 μM ) , ubiquitin ( 25 μM ) and His6-TRAF6 or His6-Tax in the presence of ATP ( 2 mM ) . After incubation , the reaction mixtures were analyzed by immunoblotting . After cell-free reactions , the reaction mixtures were separated using 6% polyacrylamide gels containing 20 μM Phos-tag acrylamide ( Wako ) and 40 μM MnCl2 . After electrophoresis , the gels were washed with transfer buffer containing 10 mM EDTA for 15 min . The gels were further washed with transfer buffer without EDTA for 10 min , and the samples were transferred to PVDF membranes , followed by immunoblotting . Statistically significant differences between mean values were determined using Student’s t-test ( **P<0 . 01 , *P<0 . 05 ) . Data are presented as the means ± SD .
|
NF-κB is a key transcription factor that regulates many physiologically important cellular processes . However , persistent activation of NF-κB leads to chronic inflammation , autoimmunity and malignancy . Infection with the human retrovirus HTLV-1 causes adult T-cell leukemia , and HTLV-1 Tax-mediated persistent NF-κB activation is crucial for leukemogenesis . Therefore , a better understanding of the precise mechanism underlying aberrant NF-κB activation is essential to develop new therapeutic approaches . Ubiquitination is one of the major post-translational modifications that regulate various intracellular signaling pathways . We and others have shown that Tax activates NF-κB through activation of the IκB kinase ( IKK ) complex by generating Lys63-linked polyubiquitin chains . However , the molecular mechanism underlying Tax-induced IKK activation remains less well understood . Here , we demonstrate precisely how HTLV-1 Tax utilizes the ubiquitin system to activate the IKK complex . The IKK complex-associated Lys63/Met1-linked hybrid polyubiquitin chains are generated through the Tax-mediated recruitment of linear ubiquitin chain assembly complex ( LUBAC ) to the IKK complex . Furthermore , the hybrid chains are required for the Tax-induced formation of the active macromolecular IKK complex . Accordingly , we propose a novel model in which Tax triggers Lys63/Met1-linked hybrid-chain-dependent oligomerization of the IKK complex , leading to trans-autophosphorylation-mediated IKK activation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"phosphorylation",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"recombination",
"reactions",
"pathology",
"and",
"laboratory",
"medicine",
"molecular",
"probe",
"techniques",
"enzymes",
"pathogens",
"gene",
"regulation",
"immunology",
"immunoblotting",
"microbiology",
"enzymology",
"retroviruses",
"viruses",
"rna",
"viruses",
"immunoprecipitation",
"molecular",
"biology",
"techniques",
"ligases",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"htlv-1",
"gene",
"expression",
"t",
"cells",
"microbial",
"pathogens",
"chemistry",
"molecular",
"biology",
"precipitation",
"techniques",
"biochemistry",
"rna",
"cell",
"biology",
"post-translational",
"modification",
"nucleic",
"acids",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences",
"cellular",
"types",
"non-coding",
"rna",
"organisms"
] |
2017
|
HTLV-1 Tax Induces Formation of the Active Macromolecular IKK Complex by Generating Lys63- and Met1-Linked Hybrid Polyubiquitin Chains
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.